diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..1edb4a97721c32c4fc13ec54febeeddd02597d77 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text +onnx/model.onnx.data filter=lfs diff=lfs merge=lfs -text +lora/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000000000000000000000000000000000000..a621daa749d03fdbee03e14c171ca746da0297b7 --- /dev/null +++ b/LICENSE @@ -0,0 +1,21 @@ +MIT License + +Copyright (c) 2026 vLLM Semantic Router contributors + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/NOTICE.md b/NOTICE.md new file mode 100644 index 0000000000000000000000000000000000000000..bcae9fcbabb381a4be526e49decd17dbab66d537 --- /dev/null +++ b/NOTICE.md @@ -0,0 +1,20 @@ +# Attribution and scope + +The new Modality task starts from a fresh adapter and prediction head on the +MIT-licensed mmBERT encoder, then adapts to the MIT-licensed Vela Base at +`5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114`. The original Modality task weights are +only an evaluation baseline; they are not a parent of this new task model. +This release's task weights are MIT licensed. + +Bundled Semantic Router training/evaluation code and original annotation/task +contract code are Apache-2.0; the full license is in `reproducibility/LICENSE`. +Global-MMLU and Aya are Apache-2.0. DiffusionDB caption metadata are CC0-1.0; +images are not included or required. MMLU-Pro is MIT and the original category +supplement is Apache-2.0; they are used for historical-source screening in the +academic-retention source construction. Pinned revisions, source licenses and +file hashes are in `reproducibility/modality-source-files.json`. + +Reviewed source labels are assistant-authored task judgments, not expert human +annotations. Authored output-contract families and background text do not become +naturally observed routing requests through their association with caption or +academic datasets. Historical encoder-pretraining exposure remains unknown. diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8f60e9ae8bcd0a770becc1f92df81aa96979851e --- /dev/null +++ b/README.md @@ -0,0 +1,228 @@ +--- +language: +- en +- zh +- es +- fr +- de +- ja +license: mit +library_name: transformers +pipeline_tag: text-classification +base_model: llm-semantic-router/Vela-1.0-Encoder-307M +tags: +- modernbert +- semantic-router +- modality-routing +- text-only +- long-context +- vela +--- + +# Vela 1.0 Encoder 307M Modality + +This **text-only** classifier identifies the requested output type: +`AR` (text/autoregressive response), `DIFFUSION` (image generation), or `BOTH` +(explicitly requested text and image outputs). It does not consume images or +audio, generate those outputs, identify an input file's media type, or choose +an arbitrary inference engine. These labels describe the router's existing +output contract, not a universal classification of generative architectures. + +The new task improves independent authored output-contract accuracy from +**34.63% to 92.96%** and long-context stress from **24/72 to 54/72** correct. +There is a regression on the small natural-request AR slice, **19/19 to 16/19**. +The authored score is not an estimate of natural production routing accuracy. + +## Model and fixed lineage + +| Property | Value | +|---|---| +| Architecture | Standard `ModernBertForSequenceClassification` | +| Shared encoder parameters | 306,939,648 | +| Complete merged task parameters | 307,532,547 | +| Pooling | Mean, explicitly serialized | +| Weights / final evaluation | FP32 | +| Maximum input capacity | 32,768 tokenizer tokens including special tokens | +| Label IDs | `AR=0`, `DIFFUSION=1`, `BOTH=2` | +| Final selected candidate | `modality-v6-vela-base`, step 200 | +| Actual final parent | `llm-semantic-router/Vela-1.0-Encoder-307M` at `5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114` | +| Merged weight SHA256 | `d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7` | + +The task began with a fresh LoRA and task head on the fixed original mmBERT +encoder, then underwent short/diverse/long curricula. The original Modality +adapter and merged task weights were not continued. After the original-parent +candidate was selected, a direct transfer to Vela Base failed development +retention; a bounded 200-step adaptation passed every predeclared gate while +preserving the trained task head. The final candidate was frozen before final +predictions. See [the freeze decision](evaluation/freeze-decision.json) and +[migration protocol](evaluation/migration-plan.json). + +The public runtime mapping is `modality_mapping.json`. The generic mapping and +HF `id2label` / `label2id` describe the same frozen IDs. Mapping files, pooling, +tokenizer and weights are locked together in +[the candidate receipt](evaluation/candidate-lock.json). + +## Independent FP32 final results + +The baseline is `llm-semantic-router/mmbert32k-modality-router-merged` at +`9f05f15889d32987d29bbe2e71bff945902ff553`. Both models receive the same full input, +fixed labels and argmax scoring with FP32 parameters/forward, without truncation. +No final outcome was used to choose weights, steps, thresholds or a new candidate. + +| Final population | Rows / source groups | Old correct | Vela correct | Old macro F1 | Vela macro F1 | +|---|---:|---:|---:|---:|---:| +| Authored multilingual output contracts with gallery captions | 1,080 / 57 | 374 (34.63%) | 1,004 (92.96%) | 0.2216 | 0.9291 | +| Academic text-answer / AR retention | 240 / 218 | 232 (96.67%) | 240 (100%) | 0.3277 | 0.3333 | +| Aya original natural requests, all AR | 19 / 6 | 19 (100%) | 16 (84.21%) | 0.3333 | 0.3048 | +| Authored long stress, EN/ZH | 72 / 6 | 24 (33.33%) | 54 (75%) | 0.2667 | 0.7395 | + +Macro F1 in the table uses all three contract labels. The academic and Aya slices +contain only AR, so a perfect result has full-contract macro F1 1/3. They test +AR retention and cannot establish DIFFUSION/BOTH performance. Aya has 15 French +rows, two Spanish, one English and one German; the new errors are two French +and the German request. These six correlated source groups provide limited +natural coverage, and their regression remains in the report. Source requests +were individually assistant-reviewed for this task, not independently annotated +by expert human raters. + +The three-way output-contract examples use held-out source groups and authored +contract families, sometimes attached to DiffusionDB captions. They are controlled +text instructions rather than naturally observed image-generation traffic. +Captions and language variants are correlated; row counts do not imply 1,080 +independent scenarios. Historical encoder-pretraining exposure is unknown. + +| Authored output-contract language | Rows | Old accuracy | Vela accuracy | +|---|---:|---:|---:| +| English | 180 | 33.33% | 97.78% | +| Chinese | 180 | 32.22% | 95.00% | +| Spanish | 180 | 35.56% | 97.22% | +| French | 180 | 33.33% | 91.11% | +| German | 180 | 38.89% | 93.33% | +| Japanese | 180 | 34.44% | 83.33% | + +Japanese remains the weakest of these authored-language slices. These language +scores are not natural-request or general multilingual understanding estimates. + +| Long stress budget | Rows | Old correct | Vela correct | Vela macro F1 | +|---|---:|---:|---:|---:| +| 4,096 | 18 | 6 | 15 | 0.8222 | +| 8,192 | 18 | 6 | 15 | 0.8222 | +| 16,384 | 18 | 6 | 13 | 0.7167 | +| 32,768 | 18 | 6 | 11 | 0.5889 | + +These 72 rows reuse six held-out source payload groups across three positions +and four lengths. Authored repeated background prose is separated from training +backgrounds, but this is a small stress test, not a natural long-document +benchmark. English improves to 32/36, Chinese to 22/36. At 32K, seven of 18 still +fail; maximum position capacity does not guarantee correct task identification. + +Complete labels, predictions, source groups, per-class/language/length scores and +paired regressions are available in +[the recomputable comparison](evaluation/comparison.json) and +[the frozen FP32 report](evaluation/final-candidate-fp32.json). + +## Data and training + +The recipe uses reviewed Aya text requests, academic text-answer examples from +Global-MMLU, authored output contracts and CC0 DiffusionDB caption metadata. +No images are downloaded. Source/duplicate groups and output-contract families +are separated across training, development and final partitions. Train-only +contract diversification adds new instruction and negative-context patterns. +Long curricula retain parent groups and combine a marked task with independent +background text; the 4K subset precedes the full 32K curriculum. + +| Stage | Maximum tokens | Steps / selected step | Learning rate | +|---|---:|---:|---:| +| Fresh short task | 2,048 | 600 / 100 | 3e-5 | +| Train-only contract diversification | 2,048 | 600 / 300 | 1e-5 | +| Diverse 4K curriculum | 4,096 | 600 / 400 | 1e-5 | +| Diverse full-context curriculum | 32,768 | 400 / 400 | 5e-6 | +| Vela Base adaptation | 32,768 | 200 / 200 | 2e-6 | + +Training uses LoRA rank 32 / alpha 64, dropout 0.05, attention/MLP projections, +complete task head and classifier, FP32 parameters, BF16 autocast, SDPA and +activation checkpointing. The 4K/full stages balance length and label; the final +adaptation keeps those sampling semantics with optional token-budget microbatch +regrouping to reduce padding. Development source-macro F1 selects the long-stage +checkpoints. Source sampling and source-weighted metrics are separate concepts. +The failed direct transfer and earlier context trial are retained as diagnostics; +training steps are not chosen from final performance. + +All 14 frozen data files were rebuilt byte-for-byte. The public initializer +reproduced the recorded initial adapter hash. The standalone package contains +source revisions/hashes/licenses, train/dev/final manifests, annotation sidecars, +training receipts, explicit export and final evaluation commands. + +## Use + +Use Transformers 4.57.6 and a compatible Torch build. FP32 is the qualified +reference precision; any additional engine/precision artifact has separate +qualification. Choose a conservative context budget for your device. + +```python +import torch +from transformers import AutoModelForSequenceClassification, AutoTokenizer + +model_id = "llm-semantic-router/Vela-1.0-Encoder-307M-Modality" +tokenizer = AutoTokenizer.from_pretrained(model_id) +model = AutoModelForSequenceClassification.from_pretrained( + model_id, torch_dtype=torch.float32, attn_implementation="sdpa" +).eval() +inputs = tokenizer("Create an illustration and explain its composition.", + return_tensors="pt", truncation=False) +if inputs["input_ids"].shape[1] > 32768: + raise ValueError("Input exceeds the model token budget") +with torch.inference_mode(): + probabilities = model(**inputs).logits.softmax(-1)[0] +label_id = int(probabilities.argmax()) +print(model.config.id2label[label_id], float(probabilities[label_id])) +``` + +Pin a repository revision for production. Softmax confidence is not calibrated +risk, and this three-label head has no explicit abstain/unknown class. Negated, +quoted or indirect image-generation instructions remain useful application-specific +validation targets. Selecting a backend, generating an image, enforcing thresholds +and deciding which part of a conversation to classify belong to the router. +Forward timings in evaluation exclude HTTP, tokenization, queueing and routing; +they are not end-to-end latency claims. + +## Reproduction and license + +Follow [the standalone guide](reproducibility/README.md) to reconstruct sources, +initialize, train, export and score. Frozen text-free labels and predictions +support independent metric recomputation. Weight hashes remain unchanged after +final evaluation. The task weights are MIT licensed; copied code is Apache-2.0 +and datasets retain their source licenses. See [NOTICE](NOTICE.md) and the +pinned source manifest for attribution and scope. + +## Qualified native and ONNX variants + +The released `onnx/model.onnx` uses FP32 query-block attention and fixes both +input batch dimensions to **one request**. Batches larger than one are explicitly +rejected. This matches the Router's single-request classifier interface; it is +not a claim that all dynamic batching modes passed. The adjacent external weight +file is required. No CK custom-op library is needed for this variant. + +On the recorded AMD ROCm execution provider, native interface comparisons cover +short inputs and 4K, 8K, 16K and 32K without truncation; inputs above 32,768 tokens +are rejected. The largest native probability/confidence difference was +0.00000042. The ONNX-level report separately checks the complete output +probability vector, unchanged argmax labels, and GPU execution of attention +inside each loop. See [native AMD evidence](evaluation/engines/vela-modality-blocked-b1-native-gpu-32k-v1.json) +and [graph/profile evidence](evaluation/engines/vela-modality-blocked-b1-onnx-gpu-v1.json). +Candle CPU numerical evidence in the same directory covers the stated short +probe set; long-context HF task-quality tables do not substitute for native +engine qualification. These are equivalence tests on fixed inputs, not another +independent accuracy test or a serving-latency benchmark. + +The `lora/` variant retains the exact development-selected adapter and pinned +parent. Use its `load_adapter.py` example to restore task configuration and +pooling before loading weights. Independent padded bilingual/Unicode CPU +comparisons against the frozen merged checkpoint are recorded in +`standalone-verification.json`; the adapter does not silently change lineage. +See [ONNX reproduction](reproducibility/onnx/README.md). + +The [storage provenance](evaluation/engines/blocked-storage-provenance.json) binds +the released graph and external data file to the qualified graph and the frozen +FP32 export. All 157 retained initializer tensors match their original dtype, +shape and bytes; graph rewriting did not alter task weights. diff --git a/adapter-qualification.json b/adapter-qualification.json new file mode 100644 index 0000000000000000000000000000000000000000..6a51dab766a95379609607c3109d6ca6244833d7 --- /dev/null +++ b/adapter-qualification.json @@ -0,0 +1,58 @@ +{ + "task": "Modality", + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "base_weight_sha256": "ccc274d12ef6f94f54c0c23f0388527b0a925268a1332b3a6e5c8e56868efb00", + "adapter_weight_sha256": "52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00", + "merged_weight_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "dtype": "float32", + "provider": "Torch CPU SDPA", + "pooling": "mean", + "explicit_frozen_task_config": true, + "probability_atol": 0.0002, + "probability_rtol": 0.0001, + "all_argmax_equal_required": true, + "quality_benchmark": false, + "long_context_qualification": false, + "results": [ + { + "batch": 2, + "padded_tokens": 14, + "attention_mask_counts": [ + 14, + 10 + ], + "input_sha256": "5b14566354399d634cd9322d69f32fbe1a3400dcc272481d96318e8370184777", + "maximum_absolute_logit_difference": 1.3828277587890625e-05, + "maximum_absolute_probability_difference": 3.259629011154175e-09, + "argmax_differences": 0, + "passed": true + }, + { + "batch": 2, + "padded_tokens": 20, + "attention_mask_counts": [ + 20, + 11 + ], + "input_sha256": "efd7d799cc51fc1a3a5dacdb84ad5267feb972e6498a7824d1d51e424b2082d2", + "maximum_absolute_logit_difference": 1.6450881958007812e-05, + "maximum_absolute_probability_difference": 2.1141022443771362e-07, + "argmax_differences": 0, + "passed": true + }, + { + "batch": 2, + "padded_tokens": 102, + "attention_mask_counts": [ + 6, + 102 + ], + "input_sha256": "3af336fced079e4eeb53e968b3bd26cbb3ed2c996b37d710c6108708ba5f8770", + "maximum_absolute_logit_difference": 8.463859558105469e-05, + "maximum_absolute_probability_difference": 2.6635825634002686e-06, + "argmax_differences": 0, + "passed": true + } + ] +} diff --git a/adapter-variant-manifest.json b/adapter-variant-manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..c9ebdec2cb7ac2608cd87b12e703f931fcac22a0 --- /dev/null +++ b/adapter-variant-manifest.json @@ -0,0 +1,19 @@ +{ + "task": "Modality", + "files": { + "adapter-qualification.json": "1c3f3675e975a08f26685ff955263f2c17e6de624115b3a627ea90ca1b5f9c98", + "lora/README.md": "46387be05f1fa1bd1fd92c09f615bf8e3bd8d9b499089f785148c1c186f0bd8e", + "lora/adapter_config.json": "426d2ca8619487565d658930462e30a0d156641f0800c255f84b3610f44139d5", + "lora/adapter_model.safetensors": "52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00", + "lora/config.json": "4cc261fd20f7acf8eb78c7c81491a57eaff5cbf5dbe73ed738fc87af97e729f2", + "lora/label_mapping.json": "63c62f883903b470c24ab5c760f593d0dbe4a01823376ee312dbf77fa9b7686a", + "lora/load_adapter.py": "759023bdc0bd563b422068c6b3dd28361a5c2ecdd3f4b8073ef931e040940298", + "lora/modality_mapping.json": "40890ec869e786e4ca6510367bd1a6cdc2feedaf8564b345b24b38bbce7e0edd", + "lora/special_tokens_map.json": "6e3204c5a4004719c185007a745d1940471a6fb02521cfa0a00306dbb6bc5903", + "lora/tokenizer.json": "22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f", + "lora/tokenizer_config.json": "74a259bb1a3811a7e3028adcd07a65765d866d5e66e0e883f0994ccfa67e8455", + "lora/verify_adapter.py": "ad9a6a32d8bc488dd56976936c4cccc217939a9c29256a069d4e9430c901b85f", + "standalone-verification.json": "1f1004e815211afa6ca714f169e311a54a6b27cc539de304f0d1064ebb6b1421" + }, + "independent_standalone_verification": true +} diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70f38b0a638eb4c674449d78ecc6f05e9ae9bd6e --- /dev/null +++ b/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ModernBertForSequenceClassification" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 2, + "classifier_activation": "gelu", + "classifier_bias": false, + "classifier_dropout": 0.0, + "classifier_pooling": "mean", + "cls_token_id": 1, + "decoder_bias": true, + "deterministic_flash_attn": false, + "dtype": "float32", + "embedding_dropout": 0.0, + "eos_token_id": 1, + "global_attn_every_n_layers": 3, + "global_rope_theta": 160000, + "gradient_checkpointing": false, + "hidden_activation": "gelu", + "hidden_size": 768, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "initializer_cutoff_factor": 2.0, + "initializer_range": 0.02, + "intermediate_size": 1152, + "label2id": { + "AR": 0, + "BOTH": 2, + "DIFFUSION": 1 + }, + "layer_norm_eps": 1e-05, + "local_attention": 128, + "local_rope_theta": 160000, + "mask_token_id": 4, + "max_position_embeddings": 32768, + "mlp_bias": false, + "mlp_dropout": 0.0, + "model_type": "modernbert", + "norm_bias": false, + "norm_eps": 1e-05, + "num_attention_heads": 12, + "num_hidden_layers": 22, + "pad_token_id": 0, + "position_embedding_type": "sans_pos", + "problem_type": "single_label_classification", + "repad_logits_with_grad": false, + "sep_token_id": 1, + "sparse_pred_ignore_index": -100, + "sparse_prediction": false, + "transformers_version": "4.57.6", + "vocab_size": 256000 +} diff --git a/evaluation/candidate-config.json b/evaluation/candidate-config.json new file mode 100644 index 0000000000000000000000000000000000000000..70f38b0a638eb4c674449d78ecc6f05e9ae9bd6e --- /dev/null +++ b/evaluation/candidate-config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ModernBertForSequenceClassification" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 2, + "classifier_activation": "gelu", + "classifier_bias": false, + "classifier_dropout": 0.0, + "classifier_pooling": "mean", + "cls_token_id": 1, + "decoder_bias": true, + "deterministic_flash_attn": false, + "dtype": "float32", + "embedding_dropout": 0.0, + "eos_token_id": 1, + "global_attn_every_n_layers": 3, + "global_rope_theta": 160000, + "gradient_checkpointing": false, + "hidden_activation": "gelu", + "hidden_size": 768, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "initializer_cutoff_factor": 2.0, + "initializer_range": 0.02, + "intermediate_size": 1152, + "label2id": { + "AR": 0, + "BOTH": 2, + "DIFFUSION": 1 + }, + "layer_norm_eps": 1e-05, + "local_attention": 128, + "local_rope_theta": 160000, + "mask_token_id": 4, + "max_position_embeddings": 32768, + "mlp_bias": false, + "mlp_dropout": 0.0, + "model_type": "modernbert", + "norm_bias": false, + "norm_eps": 1e-05, + "num_attention_heads": 12, + "num_hidden_layers": 22, + "pad_token_id": 0, + "position_embedding_type": "sans_pos", + "problem_type": "single_label_classification", + "repad_logits_with_grad": false, + "sep_token_id": 1, + "sparse_pred_ignore_index": -100, + "sparse_prediction": false, + "transformers_version": "4.57.6", + "vocab_size": 256000 +} diff --git a/evaluation/candidate-label_mapping.json b/evaluation/candidate-label_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..ec64224bd1d6536f0b6cd34ef9200616b4f3f8bd --- /dev/null +++ b/evaluation/candidate-label_mapping.json @@ -0,0 +1,22 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/evaluation/candidate-lock.json b/evaluation/candidate-lock.json new file mode 100644 index 0000000000000000000000000000000000000000..b8d290730d15633b7adc8c8145be26418766d414 --- /dev/null +++ b/evaluation/candidate-lock.json @@ -0,0 +1,34 @@ +{ + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "adapter_sha256": "52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00", + "run_manifest_sha256": "a6e0136fa8a08b0e03dd581518c2cda1d1a3468c5e32d84be0690888b350baf8", + "parameter_count": 307532547, + "encoder_parameter_count": 306939648, + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "runtime_task": "modality", + "runtime_mapping_files": [ + "label_mapping.json", + "modality_mapping.json" + ], + "task": "single_label_classification", + "classifier_pooling": "mean", + "max_position_embeddings": 32768, + "position_capacity_is_quality_evidence": false, + "weight_dtype": "float32", + "test_used_for_selection": false, + "merge_probe_max_absolute_difference": 3.230571746826172e-05, + "files": { + "config.json": "4cc261fd20f7acf8eb78c7c81491a57eaff5cbf5dbe73ed738fc87af97e729f2", + "label_mapping.json": "63c62f883903b470c24ab5c760f593d0dbe4a01823376ee312dbf77fa9b7686a", + "modality_mapping.json": "40890ec869e786e4ca6510367bd1a6cdc2feedaf8564b345b24b38bbce7e0edd", + "model.safetensors": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "special_tokens_map.json": "6e3204c5a4004719c185007a745d1940471a6fb02521cfa0a00306dbb6bc5903", + "tokenizer.json": "22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f", + "tokenizer_config.json": "74a259bb1a3811a7e3028adcd07a65765d866d5e66e0e883f0994ccfa67e8455" + } +} diff --git a/evaluation/candidate-modality_mapping.json b/evaluation/candidate-modality_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..db14d98cd37b981d8a8d1b27770c3fed99021768 --- /dev/null +++ b/evaluation/candidate-modality_mapping.json @@ -0,0 +1,12 @@ +{ + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/evaluation/comparison.json b/evaluation/comparison.json new file mode 100644 index 0000000000000000000000000000000000000000..6931b281dae8c97d29e55bd59a8c6eb5cbb4ba32 --- /dev/null +++ b/evaluation/comparison.json @@ -0,0 +1,3276 @@ +{ + "selection_used_test": false, + "metric": "Argmax classification on FP32 frozen predictions; academic and authored stress are not natural-request estimates", + "inputs": { + "manifest": "2bf2b4787248172f9df19fdc2203b20e5c83f1967f355922e5e9ba5025064fe0", + "baseline": "2fe80dcebad0334c54c8d1bd064346d2e9e702a00fb01fba4b7186cd518d8c4e", + "candidate": "2f28bc2e3f7c9a6cd3898efe6210202fc75c4d265158210c1f148635cf8a131c", + "contract": "87c8b1584d114e307901a3e1a203d93433cffbea0e62e3810825009a5f280a2a" + }, + "sources": { + "academic-text-answer": { + "rows": 240, + "source_groups": 218, + "baseline": { + "rows": 240, + "accuracy": 0.9666666666666667, + "macro_f1": 0.3276836158192091, + "macro_f1_present_labels": 0.9830508474576272, + "per_label": { + "AR": { + "support": 240, + "tp": 232, + "fp": 0, + "fn": 8, + "precision": 1.0, + "recall": 0.9666666666666667, + "f1": 0.9830508474576272 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 8, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 232, + 0, + 8 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 240, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 240, + "tp": 240, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 240, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 8, + "regressed": 0 + }, + "by_language": { + "de": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 0 + } + }, + "en": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 0.975, + "macro_f1": 0.32911392405063294, + "macro_f1_present_labels": 0.9873417721518988, + "per_label": { + "AR": { + "support": 40, + "tp": 39, + "fp": 0, + "fn": 1, + "precision": 1.0, + "recall": 0.975, + "f1": 0.9873417721518988 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 39, + 0, + 1 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 1, + "regressed": 0 + } + }, + "es": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 0 + } + }, + "fr": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 0 + } + }, + "ja": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 0.9, + "macro_f1": 0.3157894736842105, + "macro_f1_present_labels": 0.9473684210526315, + "per_label": { + "AR": { + "support": 40, + "tp": 36, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.9, + "f1": 0.9473684210526315 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 4, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 36, + 0, + 4 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 4, + "regressed": 0 + } + }, + "zh": { + "rows": 40, + "source_groups": 40, + "baseline": { + "rows": 40, + "accuracy": 0.925, + "macro_f1": 0.3203463203463203, + "macro_f1_present_labels": 0.961038961038961, + "per_label": { + "AR": { + "support": 40, + "tp": 37, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.925, + "f1": 0.961038961038961 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 3, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 37, + 0, + 3 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 40, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 40, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 3, + "regressed": 0 + } + } + }, + "by_length": { + "short": { + "rows": 240, + "source_groups": 218, + "baseline": { + "rows": 240, + "accuracy": 0.9666666666666667, + "macro_f1": 0.3276836158192091, + "macro_f1_present_labels": 0.9830508474576272, + "per_label": { + "AR": { + "support": 240, + "tp": 232, + "fp": 0, + "fn": 8, + "precision": 1.0, + "recall": 0.9666666666666667, + "f1": 0.9830508474576272 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 8, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 232, + 0, + 8 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 240, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 240, + "tp": 240, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 240, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 8, + "regressed": 0 + } + } + } + }, + "authored-context-on-frozen-source": { + "rows": 72, + "source_groups": 6, + "baseline": { + "rows": 72, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 24, + "fn": 12, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 24, + "tp": 12, + "fp": 24, + "fn": 12, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 24, + "tp": 0, + "fp": 0, + "fn": 24, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 12, + 12, + 0 + ], + [ + 12, + 12, + 0 + ] + ] + }, + "candidate": { + "rows": 72, + "accuracy": 0.75, + "macro_f1": 0.7394880644233535, + "macro_f1_present_labels": 0.7394880644233535, + "per_label": { + "AR": { + "support": 24, + "tp": 18, + "fp": 3, + "fn": 6, + "precision": 0.8571428571428571, + "recall": 0.75, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 24, + "tp": 12, + "fp": 2, + "fn": 12, + "precision": 0.8571428571428571, + "recall": 0.5, + "f1": 0.631578947368421 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 13, + "fn": 0, + "precision": 0.6486486486486487, + "recall": 1.0, + "f1": 0.7868852459016393 + } + }, + "confusion": [ + [ + 18, + 2, + 4 + ], + [ + 3, + 12, + 9 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "paired_correct_changes": { + "improved": 32, + "regressed": 2 + }, + "by_language": { + "en": { + "rows": 36, + "source_groups": 3, + "baseline": { + "rows": 36, + "accuracy": 0.3333333333333333, + "macro_f1": 0.16666666666666666, + "macro_f1_present_labels": 0.16666666666666666, + "per_label": { + "AR": { + "support": 12, + "tp": 0, + "fp": 0, + "fn": 12, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "DIFFUSION": { + "support": 12, + "tp": 12, + "fp": 24, + "fn": 0, + "precision": 0.3333333333333333, + "recall": 1.0, + "f1": 0.5 + }, + "BOTH": { + "support": 12, + "tp": 0, + "fp": 0, + "fn": 12, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 0, + 12, + 0 + ], + [ + 0, + 12, + 0 + ], + [ + 0, + 12, + 0 + ] + ] + }, + "candidate": { + "rows": 36, + "accuracy": 0.8888888888888888, + "macro_f1": 0.8857142857142857, + "macro_f1_present_labels": 0.8857142857142857, + "per_label": { + "AR": { + "support": 12, + "tp": 8, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.6666666666666666, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 12, + "tp": 12, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 12, + "tp": 12, + "fp": 4, + "fn": 0, + "precision": 0.75, + "recall": 1.0, + "f1": 0.8571428571428571 + } + }, + "confusion": [ + [ + 8, + 0, + 4 + ], + [ + 0, + 12, + 0 + ], + [ + 0, + 0, + 12 + ] + ] + }, + "paired_correct_changes": { + "improved": 20, + "regressed": 0 + } + }, + "zh": { + "rows": 36, + "source_groups": 3, + "baseline": { + "rows": 36, + "accuracy": 0.3333333333333333, + "macro_f1": 0.16666666666666666, + "macro_f1_present_labels": 0.16666666666666666, + "per_label": { + "AR": { + "support": 12, + "tp": 12, + "fp": 24, + "fn": 0, + "precision": 0.3333333333333333, + "recall": 1.0, + "f1": 0.5 + }, + "DIFFUSION": { + "support": 12, + "tp": 0, + "fp": 0, + "fn": 12, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 12, + "tp": 0, + "fp": 0, + "fn": 12, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 12, + 0, + 0 + ], + [ + 12, + 0, + 0 + ], + [ + 12, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 36, + "accuracy": 0.6111111111111112, + "macro_f1": 0.5090909090909091, + "macro_f1_present_labels": 0.5090909090909091, + "per_label": { + "AR": { + "support": 12, + "tp": 10, + "fp": 3, + "fn": 2, + "precision": 0.7692307692307693, + "recall": 0.8333333333333334, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 12, + "tp": 0, + "fp": 2, + "fn": 12, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 12, + "tp": 12, + "fp": 9, + "fn": 0, + "precision": 0.5714285714285714, + "recall": 1.0, + "f1": 0.7272727272727273 + } + }, + "confusion": [ + [ + 10, + 2, + 0 + ], + [ + 3, + 0, + 9 + ], + [ + 0, + 0, + 12 + ] + ] + }, + "paired_correct_changes": { + "improved": 12, + "regressed": 2 + } + } + }, + "by_length": { + "16384": { + "rows": 18, + "source_groups": 6, + "baseline": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "candidate": { + "rows": 18, + "accuracy": 0.7222222222222222, + "macro_f1": 0.7166666666666667, + "macro_f1_present_labels": 0.7166666666666667, + "per_label": { + "AR": { + "support": 6, + "tp": 4, + "fp": 0, + "fn": 2, + "precision": 1.0, + "recall": 0.6666666666666666, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 1, + "fn": 3, + "precision": 0.75, + "recall": 0.5, + "f1": 0.6 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 4, + "fn": 0, + "precision": 0.6, + "recall": 1.0, + "f1": 0.75 + } + }, + "confusion": [ + [ + 4, + 1, + 1 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "paired_correct_changes": { + "improved": 8, + "regressed": 1 + } + }, + "32768": { + "rows": 18, + "source_groups": 6, + "baseline": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "candidate": { + "rows": 18, + "accuracy": 0.6111111111111112, + "macro_f1": 0.5888888888888889, + "macro_f1_present_labels": 0.5888888888888889, + "per_label": { + "AR": { + "support": 6, + "tp": 2, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.3333333333333333, + "f1": 0.5 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 1, + "fn": 3, + "precision": 0.75, + "recall": 0.5, + "f1": 0.6 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 6, + "fn": 0, + "precision": 0.5, + "recall": 1.0, + "f1": 0.6666666666666666 + } + }, + "confusion": [ + [ + 2, + 1, + 3 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "paired_correct_changes": { + "improved": 6, + "regressed": 1 + } + }, + "4096": { + "rows": 18, + "source_groups": 6, + "baseline": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "candidate": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 6, + 0, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "paired_correct_changes": { + "improved": 9, + "regressed": 0 + } + }, + "8192": { + "rows": 18, + "source_groups": 6, + "baseline": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "candidate": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + } + }, + "confusion": [ + [ + 6, + 0, + 0 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "paired_correct_changes": { + "improved": 9, + "regressed": 0 + } + } + } + }, + "authored-output-contract-with-gallery-caption": { + "rows": 1080, + "source_groups": 57, + "baseline": { + "rows": 1080, + "accuracy": 0.34629629629629627, + "macro_f1": 0.2216396874882791, + "macro_f1_present_labels": 0.2216396874882791, + "per_label": { + "AR": { + "support": 360, + "tp": 31, + "fp": 46, + "fn": 329, + "precision": 0.4025974025974026, + "recall": 0.08611111111111111, + "f1": 0.14187643020594964 + }, + "DIFFUSION": { + "support": 360, + "tp": 339, + "fp": 653, + "fn": 21, + "precision": 0.34173387096774194, + "recall": 0.9416666666666667, + "f1": 0.5014792899408284 + }, + "BOTH": { + "support": 360, + "tp": 4, + "fp": 7, + "fn": 356, + "precision": 0.36363636363636365, + "recall": 0.011111111111111112, + "f1": 0.0215633423180593 + } + }, + "confusion": [ + [ + 31, + 323, + 6 + ], + [ + 20, + 339, + 1 + ], + [ + 26, + 330, + 4 + ] + ] + }, + "candidate": { + "rows": 1080, + "accuracy": 0.9296296296296296, + "macro_f1": 0.9290630636363364, + "macro_f1_present_labels": 0.9290630636363364, + "per_label": { + "AR": { + "support": 360, + "tp": 358, + "fp": 70, + "fn": 2, + "precision": 0.8364485981308412, + "recall": 0.9944444444444445, + "f1": 0.9086294416243654 + }, + "DIFFUSION": { + "support": 360, + "tp": 290, + "fp": 6, + "fn": 70, + "precision": 0.9797297297297297, + "recall": 0.8055555555555556, + "f1": 0.8841463414634146 + }, + "BOTH": { + "support": 360, + "tp": 356, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.9888888888888889, + "f1": 0.994413407821229 + } + }, + "confusion": [ + [ + 358, + 2, + 0 + ], + [ + 70, + 290, + 0 + ], + [ + 0, + 4, + 356 + ] + ] + }, + "paired_correct_changes": { + "improved": 692, + "regressed": 62 + }, + "by_language": { + "de": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.3888888888888889, + "macro_f1": 0.2765939238339752, + "macro_f1_present_labels": 0.2765939238339752, + "per_label": { + "AR": { + "support": 60, + "tp": 13, + "fp": 22, + "fn": 47, + "precision": 0.37142857142857144, + "recall": 0.21666666666666667, + "f1": 0.2736842105263158 + }, + "DIFFUSION": { + "support": 60, + "tp": 57, + "fp": 88, + "fn": 3, + "precision": 0.3931034482758621, + "recall": 0.95, + "f1": 0.5560975609756098 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 13, + 47, + 0 + ], + [ + 3, + 57, + 0 + ], + [ + 19, + 41, + 0 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.9333333333333333, + "macro_f1": 0.9326599326599326, + "macro_f1_present_labels": 0.9326599326599326, + "per_label": { + "AR": { + "support": 60, + "tp": 60, + "fp": 12, + "fn": 0, + "precision": 0.8333333333333334, + "recall": 1.0, + "f1": 0.9090909090909091 + }, + "DIFFUSION": { + "support": 60, + "tp": 48, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.8, + "f1": 0.8888888888888888 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 60, + 0, + 0 + ], + [ + 12, + 48, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "paired_correct_changes": { + "improved": 109, + "regressed": 11 + } + }, + "en": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.3333333333333333, + "macro_f1": 0.16666666666666666, + "macro_f1_present_labels": 0.16666666666666666, + "per_label": { + "AR": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "DIFFUSION": { + "support": 60, + "tp": 60, + "fp": 120, + "fn": 0, + "precision": 0.3333333333333333, + "recall": 1.0, + "f1": 0.5 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 0, + 60, + 0 + ], + [ + 0, + 60, + 0 + ], + [ + 0, + 60, + 0 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.9777777777777777, + "macro_f1": 0.9777716032231175, + "macro_f1_present_labels": 0.9777716032231175, + "per_label": { + "AR": { + "support": 60, + "tp": 59, + "fp": 3, + "fn": 1, + "precision": 0.9516129032258065, + "recall": 0.9833333333333333, + "f1": 0.9672131147540983 + }, + "DIFFUSION": { + "support": 60, + "tp": 57, + "fp": 1, + "fn": 3, + "precision": 0.9827586206896551, + "recall": 0.95, + "f1": 0.9661016949152542 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 59, + 1, + 0 + ], + [ + 3, + 57, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "paired_correct_changes": { + "improved": 119, + "regressed": 3 + } + }, + "es": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.35555555555555557, + "macro_f1": 0.22549019607843138, + "macro_f1_present_labels": 0.22549019607843138, + "per_label": { + "AR": { + "support": 60, + "tp": 6, + "fp": 2, + "fn": 54, + "precision": 0.75, + "recall": 0.1, + "f1": 0.17647058823529413 + }, + "DIFFUSION": { + "support": 60, + "tp": 58, + "fp": 114, + "fn": 2, + "precision": 0.3372093023255814, + "recall": 0.9666666666666667, + "f1": 0.5 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 6, + 54, + 0 + ], + [ + 2, + 58, + 0 + ], + [ + 0, + 60, + 0 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.9722222222222222, + "macro_f1": 0.972214405895479, + "macro_f1_present_labels": 0.972214405895479, + "per_label": { + "AR": { + "support": 60, + "tp": 60, + "fp": 2, + "fn": 0, + "precision": 0.967741935483871, + "recall": 1.0, + "f1": 0.9836065573770492 + }, + "DIFFUSION": { + "support": 60, + "tp": 58, + "fp": 3, + "fn": 2, + "precision": 0.9508196721311475, + "recall": 0.9666666666666667, + "f1": 0.9586776859504132 + }, + "BOTH": { + "support": 60, + "tp": 57, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.95, + "f1": 0.9743589743589743 + } + }, + "confusion": [ + [ + 60, + 0, + 0 + ], + [ + 2, + 58, + 0 + ], + [ + 0, + 3, + 57 + ] + ] + }, + "paired_correct_changes": { + "improved": 113, + "regressed": 2 + } + }, + "fr": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.3333333333333333, + "macro_f1": 0.17819217819217817, + "macro_f1_present_labels": 0.17819217819217817, + "per_label": { + "AR": { + "support": 60, + "tp": 1, + "fp": 5, + "fn": 59, + "precision": 0.16666666666666666, + "recall": 0.016666666666666666, + "f1": 0.030303030303030304 + }, + "DIFFUSION": { + "support": 60, + "tp": 59, + "fp": 115, + "fn": 1, + "precision": 0.3390804597701149, + "recall": 0.9833333333333333, + "f1": 0.5042735042735043 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 59, + 0 + ], + [ + 1, + 59, + 0 + ], + [ + 4, + 56, + 0 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.9111111111111111, + "macro_f1": 0.9098473771059786, + "macro_f1_present_labels": 0.9098473771059786, + "per_label": { + "AR": { + "support": 60, + "tp": 60, + "fp": 15, + "fn": 0, + "precision": 0.8, + "recall": 1.0, + "f1": 0.8888888888888888 + }, + "DIFFUSION": { + "support": 60, + "tp": 45, + "fp": 1, + "fn": 15, + "precision": 0.9782608695652174, + "recall": 0.75, + "f1": 0.8490566037735849 + }, + "BOTH": { + "support": 60, + "tp": 59, + "fp": 0, + "fn": 1, + "precision": 1.0, + "recall": 0.9833333333333333, + "f1": 0.9915966386554622 + } + }, + "confusion": [ + [ + 60, + 0, + 0 + ], + [ + 15, + 45, + 0 + ], + [ + 0, + 1, + 59 + ] + ] + }, + "paired_correct_changes": { + "improved": 119, + "regressed": 15 + } + }, + "ja": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.34444444444444444, + "macro_f1": 0.26517856622425845, + "macro_f1_present_labels": 0.26517856622425845, + "per_label": { + "AR": { + "support": 60, + "tp": 10, + "fp": 13, + "fn": 50, + "precision": 0.43478260869565216, + "recall": 0.16666666666666666, + "f1": 0.24096385542168675 + }, + "DIFFUSION": { + "support": 60, + "tp": 49, + "fp": 103, + "fn": 11, + "precision": 0.3223684210526316, + "recall": 0.8166666666666667, + "f1": 0.46226415094339623 + }, + "BOTH": { + "support": 60, + "tp": 3, + "fp": 2, + "fn": 57, + "precision": 0.6, + "recall": 0.05, + "f1": 0.09230769230769231 + } + }, + "confusion": [ + [ + 10, + 48, + 2 + ], + [ + 11, + 49, + 0 + ], + [ + 2, + 55, + 3 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 60, + "tp": 60, + "fp": 30, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 60, + "tp": 30, + "fp": 0, + "fn": 30, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 60, + 0, + 0 + ], + [ + 30, + 30, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "paired_correct_changes": { + "improved": 111, + "regressed": 23 + } + }, + "zh": { + "rows": 180, + "source_groups": 57, + "baseline": { + "rows": 180, + "accuracy": 0.32222222222222224, + "macro_f1": 0.18338507683485847, + "macro_f1_present_labels": 0.18338507683485847, + "per_label": { + "AR": { + "support": 60, + "tp": 1, + "fp": 4, + "fn": 59, + "precision": 0.2, + "recall": 0.016666666666666666, + "f1": 0.03076923076923077 + }, + "DIFFUSION": { + "support": 60, + "tp": 56, + "fp": 113, + "fn": 4, + "precision": 0.33136094674556216, + "recall": 0.9333333333333333, + "f1": 0.4890829694323144 + }, + "BOTH": { + "support": 60, + "tp": 1, + "fp": 5, + "fn": 59, + "precision": 0.16666666666666666, + "recall": 0.016666666666666666, + "f1": 0.030303030303030304 + } + }, + "confusion": [ + [ + 1, + 55, + 4 + ], + [ + 3, + 56, + 1 + ], + [ + 1, + 58, + 1 + ] + ] + }, + "candidate": { + "rows": 180, + "accuracy": 0.95, + "macro_f1": 0.9498292801895337, + "macro_f1_present_labels": 0.9498292801895337, + "per_label": { + "AR": { + "support": 60, + "tp": 59, + "fp": 8, + "fn": 1, + "precision": 0.8805970149253731, + "recall": 0.9833333333333333, + "f1": 0.9291338582677166 + }, + "DIFFUSION": { + "support": 60, + "tp": 52, + "fp": 1, + "fn": 8, + "precision": 0.9811320754716981, + "recall": 0.8666666666666667, + "f1": 0.9203539823008849 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 59, + 1, + 0 + ], + [ + 8, + 52, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "paired_correct_changes": { + "improved": 121, + "regressed": 8 + } + } + }, + "by_length": { + "short": { + "rows": 1080, + "source_groups": 57, + "baseline": { + "rows": 1080, + "accuracy": 0.34629629629629627, + "macro_f1": 0.2216396874882791, + "macro_f1_present_labels": 0.2216396874882791, + "per_label": { + "AR": { + "support": 360, + "tp": 31, + "fp": 46, + "fn": 329, + "precision": 0.4025974025974026, + "recall": 0.08611111111111111, + "f1": 0.14187643020594964 + }, + "DIFFUSION": { + "support": 360, + "tp": 339, + "fp": 653, + "fn": 21, + "precision": 0.34173387096774194, + "recall": 0.9416666666666667, + "f1": 0.5014792899408284 + }, + "BOTH": { + "support": 360, + "tp": 4, + "fp": 7, + "fn": 356, + "precision": 0.36363636363636365, + "recall": 0.011111111111111112, + "f1": 0.0215633423180593 + } + }, + "confusion": [ + [ + 31, + 323, + 6 + ], + [ + 20, + 339, + 1 + ], + [ + 26, + 330, + 4 + ] + ] + }, + "candidate": { + "rows": 1080, + "accuracy": 0.9296296296296296, + "macro_f1": 0.9290630636363364, + "macro_f1_present_labels": 0.9290630636363364, + "per_label": { + "AR": { + "support": 360, + "tp": 358, + "fp": 70, + "fn": 2, + "precision": 0.8364485981308412, + "recall": 0.9944444444444445, + "f1": 0.9086294416243654 + }, + "DIFFUSION": { + "support": 360, + "tp": 290, + "fp": 6, + "fn": 70, + "precision": 0.9797297297297297, + "recall": 0.8055555555555556, + "f1": 0.8841463414634146 + }, + "BOTH": { + "support": 360, + "tp": 356, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.9888888888888889, + "f1": 0.994413407821229 + } + }, + "confusion": [ + [ + 358, + 2, + 0 + ], + [ + 70, + 290, + 0 + ], + [ + 0, + 4, + 356 + ] + ] + }, + "paired_correct_changes": { + "improved": 692, + "regressed": 62 + } + } + } + }, + "aya-original-human": { + "rows": 19, + "source_groups": 6, + "baseline": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 19, + "accuracy": 0.8421052631578947, + "macro_f1": 0.30476190476190473, + "macro_f1_present_labels": 0.9142857142857143, + "per_label": { + "AR": { + "support": 19, + "tp": 16, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.8421052631578947, + "f1": 0.9142857142857143 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 2, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 16, + 1, + 2 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 3 + }, + "by_language": { + "de": { + "rows": 1, + "source_groups": 1, + "baseline": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 1, + "accuracy": 0.0, + "macro_f1": 0.0, + "macro_f1_present_labels": 0.0, + "per_label": { + "AR": { + "support": 1, + "tp": 0, + "fp": 0, + "fn": 1, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 0, + 0, + 1 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 1 + } + }, + "en": { + "rows": 1, + "source_groups": 1, + "baseline": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 0 + } + }, + "es": { + "rows": 2, + "source_groups": 2, + "baseline": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 0 + } + }, + "fr": { + "rows": 15, + "source_groups": 2, + "baseline": { + "rows": 15, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 15, + "tp": 15, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 15, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 15, + "accuracy": 0.8666666666666667, + "macro_f1": 0.30952380952380953, + "macro_f1_present_labels": 0.9285714285714286, + "per_label": { + "AR": { + "support": 15, + "tp": 13, + "fp": 0, + "fn": 2, + "precision": 1.0, + "recall": 0.8666666666666667, + "f1": 0.9285714285714286 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 13, + 1, + 1 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 2 + } + } + }, + "by_length": { + "short": { + "rows": 19, + "source_groups": 6, + "baseline": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "candidate": { + "rows": 19, + "accuracy": 0.8421052631578947, + "macro_f1": 0.30476190476190473, + "macro_f1_present_labels": 0.9142857142857143, + "per_label": { + "AR": { + "support": 19, + "tp": 16, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.8421052631578947, + "f1": 0.9142857142857143 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 2, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 16, + 1, + 2 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "paired_correct_changes": { + "improved": 0, + "regressed": 3 + } + } + } + } + } +} diff --git a/evaluation/engines/blocked-storage-provenance.json b/evaluation/engines/blocked-storage-provenance.json new file mode 100644 index 0000000000000000000000000000000000000000..91fd898b32b83c5302f456a6aa3cdca2dbc0dba6 --- /dev/null +++ b/evaluation/engines/blocked-storage-provenance.json @@ -0,0 +1,3850 @@ +{ + "schema_version": 1, + "task": "Modality", + "model": "Vela-1.0-Encoder-307M-Modality", + "passed": true, + "method": "CPU streaming SHA256 of original and final external initializer byte ranges, with exact dtype and shape equality; inline initializers compared by logical numeric bytes. Original files checked against frozen export receipt. Qualified final graph checked against historical GPU receipt. No inference rerun.", + "scope": "This receipt binds current external data to the originally exported frozen model weights and the already qualified graph. It adds storage provenance; it does not change or rerun accuracy or numerical engine qualification.", + "native_weights_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "original_export_receipt_sha256": "50f867db7e2155acf677b797468f8f0513a4cdaa7a7986e51296d5bdfb6525aa", + "original_files": { + "model.onnx": "fa09cedb1da477357a74bda382942f0ce0e341a44cde20f46d48bfc0e94af836", + "model.onnx.data": "c4c8565db328f68fda33b64550010813880a32b26cb1988e509acc979303bab8" + }, + "qualified_profile_sha256": "7264fa901a81c6078cd8cabb825d38b18a39746b19eff3ecc692813a256ec1c6", + "final_files": { + "model.onnx": "74a08c52ce47f522071fd71acee080c0a3a74ad12e3d5fb19ae1dfd244f1059c", + "model.onnx.data": "f0372b44dd9a9ae0d93f0ee0211ec27644dd2ad94351c8022e94c953bd14c399" + }, + "original_initializer_count": 161, + "final_initializer_count": 157, + "matched_external_initializers": 137, + "matched_inline_initializers": 20, + "matched_tensor_bytes": 1230130444, + "all_initializers_equal": true, + "initializers": [ + { + "name": "model.model.embeddings.tok_embeddings.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 256000, + 768 + ], + "bytes": 786432000, + "sha256": "8d7917c0766aca15106eb571153b4c0da024d499d4e8d57413957db59edfefd4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 256000, + 768 + ], + "bytes": 786432000, + "sha256": "8d7917c0766aca15106eb571153b4c0da024d499d4e8d57413957db59edfefd4" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.embeddings.norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "c48411993b1c6bfad0f7c0994958526e9d41a4c439369fb9ccabf1b069992801" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "c48411993b1c6bfad0f7c0994958526e9d41a4c439369fb9ccabf1b069992801" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.0.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ec89f8bd25efe9bee80b2b311322394efe73b8b343ddc8e862b94385a1158294" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ec89f8bd25efe9bee80b2b311322394efe73b8b343ddc8e862b94385a1158294" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.1.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "4f165660f05290a797d077ed63f44be146c7edc7a2e20a16ccdedbcfc5fa1a8b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "4f165660f05290a797d077ed63f44be146c7edc7a2e20a16ccdedbcfc5fa1a8b" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.1.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e2b7b7d3d1ce0de28043c771d0df95dbe69985d6063fc517a7df2e03a8fb5e3a" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e2b7b7d3d1ce0de28043c771d0df95dbe69985d6063fc517a7df2e03a8fb5e3a" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.2.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "b923f7ec46d4e08ae4e1da80e8378ed3d66cf82e4d5773f1567bb113d9d86913" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "b923f7ec46d4e08ae4e1da80e8378ed3d66cf82e4d5773f1567bb113d9d86913" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.2.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "88d7dddd88cad01cbe1c51c3eadb58e9798d854cf09d5892ec51d3c486955e54" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "88d7dddd88cad01cbe1c51c3eadb58e9798d854cf09d5892ec51d3c486955e54" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.3.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "23675a4ac01236e2c4b0e9b2c8821be1140500619962b3c2e48b809ec142ad2e" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "23675a4ac01236e2c4b0e9b2c8821be1140500619962b3c2e48b809ec142ad2e" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.3.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "f2179bdf7f369fcd8c79493e2da8f5cb069cff6f5c8d7da0613b2a33c512791f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "f2179bdf7f369fcd8c79493e2da8f5cb069cff6f5c8d7da0613b2a33c512791f" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.4.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "eb8b281258b9a7788a5c5a903de6a088b032f5994a6f236dab9d0d945f06bdc5" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "eb8b281258b9a7788a5c5a903de6a088b032f5994a6f236dab9d0d945f06bdc5" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.4.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "a9e7de99faf4a3e66871e0510128390e03ceebfc3da5170bbe6b5e3f17ab7754" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "a9e7de99faf4a3e66871e0510128390e03ceebfc3da5170bbe6b5e3f17ab7754" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.5.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "6165f78c54ce4cf69bc769f70bba15c697ad3d6589ddc7ff7fb6d62a84e44271" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "6165f78c54ce4cf69bc769f70bba15c697ad3d6589ddc7ff7fb6d62a84e44271" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.5.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5b2447b0cbf5acd5ba44f63763232bb2e71e3eebbaffc952fb831024d8608fda" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5b2447b0cbf5acd5ba44f63763232bb2e71e3eebbaffc952fb831024d8608fda" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.6.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "f853dd2ca0d74178d689dd9981d30649af944e84e2a21dbacd06fb314cec6e86" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "f853dd2ca0d74178d689dd9981d30649af944e84e2a21dbacd06fb314cec6e86" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.6.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "049b02dd2dab0ee620f3bccf6a0f29c3b0dbc75cf5d1cc9822c7c0db867cfa3e" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "049b02dd2dab0ee620f3bccf6a0f29c3b0dbc75cf5d1cc9822c7c0db867cfa3e" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.7.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "299fe6ec9c2146dc027b62c7fd10cb656de62f5ea751754b1377585ec9ad79bb" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "299fe6ec9c2146dc027b62c7fd10cb656de62f5ea751754b1377585ec9ad79bb" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.7.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "85913a398c2060a9542b950c743930d96d1135a3aa2a07c5ac5320f127bba7b1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "85913a398c2060a9542b950c743930d96d1135a3aa2a07c5ac5320f127bba7b1" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.8.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e5ade1a4f393518288d512258a4dadd7b67410305937d685df26721bb6f83671" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e5ade1a4f393518288d512258a4dadd7b67410305937d685df26721bb6f83671" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.8.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "72c4d8fdfbc03a07e1b17886ac6c7b58848dddb49e4f27966fdcb6b2b92114e1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "72c4d8fdfbc03a07e1b17886ac6c7b58848dddb49e4f27966fdcb6b2b92114e1" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.9.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "01083bbdc4d6d22d3ac502f52a9004cb2c2792c10d3ebb4cad5a78fa2c7ccd86" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "01083bbdc4d6d22d3ac502f52a9004cb2c2792c10d3ebb4cad5a78fa2c7ccd86" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.9.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "6128518865106554d16cc194ed9b3db56a036eb07335f7ad2e13f07431a56ab2" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "6128518865106554d16cc194ed9b3db56a036eb07335f7ad2e13f07431a56ab2" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.10.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "0a0118ef2b658d932b1311c6dee628b15d9a158ffe58c805178b13c0f888cd12" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "0a0118ef2b658d932b1311c6dee628b15d9a158ffe58c805178b13c0f888cd12" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.10.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "74f241bb0dae5238ac7cf47ba501c8e0331ae68cfad85c218f361a0d776e8726" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "74f241bb0dae5238ac7cf47ba501c8e0331ae68cfad85c218f361a0d776e8726" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.11.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "70ef5e9993bd7164cfaf4ea451f23c642b74a5ea23d4b0a572bd85a4114ed6e9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "70ef5e9993bd7164cfaf4ea451f23c642b74a5ea23d4b0a572bd85a4114ed6e9" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.11.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ecfd4c6b51090624056d99d88c304520c0bdf4def4ea5d2901c07060cec73dd1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ecfd4c6b51090624056d99d88c304520c0bdf4def4ea5d2901c07060cec73dd1" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.12.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e3d47393ffffd762884b97d853be0d8899f0869d426e86556ba07437729ffdb5" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e3d47393ffffd762884b97d853be0d8899f0869d426e86556ba07437729ffdb5" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.12.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "218e71aec18c8b78c5103c9dbf959912a237fad41abd71e22978c6c758216f84" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "218e71aec18c8b78c5103c9dbf959912a237fad41abd71e22978c6c758216f84" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.13.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "76ac261daa6058a20f83af3abf067366a989b9018e1755233290009f0ed579b8" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "76ac261daa6058a20f83af3abf067366a989b9018e1755233290009f0ed579b8" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.13.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "67953b22d5373d3825ad2c10307df9dba38a9c2a68ea89e0cea0255a390233d8" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "67953b22d5373d3825ad2c10307df9dba38a9c2a68ea89e0cea0255a390233d8" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.14.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "bef605095795dfe6d75632489cd9d7bf3c378e639cb29a840d60dd3391018c8f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "bef605095795dfe6d75632489cd9d7bf3c378e639cb29a840d60dd3391018c8f" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.14.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "0de4044b70d8c3b7e9d21dd1a27b5e341846cdf22a16e426654cb5cb3ae83207" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "0de4044b70d8c3b7e9d21dd1a27b5e341846cdf22a16e426654cb5cb3ae83207" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.15.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "61121db4134fef26eff12388a48b29bbcf5d661626312f7ee9306fcc3bddbecd" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "61121db4134fef26eff12388a48b29bbcf5d661626312f7ee9306fcc3bddbecd" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.15.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e4557d75724bda343f92103ad32f1489b07748e90372aadf079b0381ae3358eb" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "e4557d75724bda343f92103ad32f1489b07748e90372aadf079b0381ae3358eb" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.16.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "42703d3fc95410028eeefecc489358196e04f6a9638adeab4b1414207082dd11" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "42703d3fc95410028eeefecc489358196e04f6a9638adeab4b1414207082dd11" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.16.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5769e4a44c4b44633b189835fec47e54af362ee6b6a822d413d5ea04b77054eb" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5769e4a44c4b44633b189835fec47e54af362ee6b6a822d413d5ea04b77054eb" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.17.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "bdb5ce1ff0abc1e9eaf6de8abb739ca494ed1903ea7bb45b3c4ef8d16a4cfa27" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "bdb5ce1ff0abc1e9eaf6de8abb739ca494ed1903ea7bb45b3c4ef8d16a4cfa27" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.17.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5821d7cd3931480d2e6c9c635df7eb3ebda3759c3ad215f2ce5a07ad47703566" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5821d7cd3931480d2e6c9c635df7eb3ebda3759c3ad215f2ce5a07ad47703566" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.18.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "da985d47d37e58ed52797efd01955c2946cc90512959ae100510424ab6590f2c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "da985d47d37e58ed52797efd01955c2946cc90512959ae100510424ab6590f2c" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.18.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5823e60a0c9c797c545d0681928b2efc6daccbdb3f5e2271bb385458a9d6c47a" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5823e60a0c9c797c545d0681928b2efc6daccbdb3f5e2271bb385458a9d6c47a" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.19.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "d34e3a181c091e7d51514bd7546ce2aa0a707ff2a23e8179fabe6f13922e0233" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "d34e3a181c091e7d51514bd7546ce2aa0a707ff2a23e8179fabe6f13922e0233" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.19.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "922ba23b773990b3ecb180b92c47d03b8fa8f2568c08a121c44b6cf5e7cc9325" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "922ba23b773990b3ecb180b92c47d03b8fa8f2568c08a121c44b6cf5e7cc9325" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.20.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "cf6c5ac9804e99c73381a8322a05624d0a22f3040eb4693d0cf9244d9d77f7b0" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "cf6c5ac9804e99c73381a8322a05624d0a22f3040eb4693d0cf9244d9d77f7b0" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.20.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5e77e8f62d46c7d9e09f3563a10bde72181f6b4539c61aedb7ce6212350eb5a7" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "5e77e8f62d46c7d9e09f3563a10bde72181f6b4539c61aedb7ce6212350eb5a7" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.21.attn_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "b15986db9b65b65b9ce0967384ed1282ef590042a16d67f74767db55c0e5458d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "b15986db9b65b65b9ce0967384ed1282ef590042a16d67f74767db55c0e5458d" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.layers.21.mlp_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ca3cbd3aebdc154c44c64a11a368a7b0755dccd31fc6777880db59e6596fafca" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "ca3cbd3aebdc154c44c64a11a368a7b0755dccd31fc6777880db59e6596fafca" + }, + "external": true, + "passed": true + }, + { + "name": "model.model.final_norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "02803ade53a2d5463eaf5352e7d798a45ce935444dae0a50d9eb97db9d0a77d1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "02803ade53a2d5463eaf5352e7d798a45ce935444dae0a50d9eb97db9d0a77d1" + }, + "external": true, + "passed": true + }, + { + "name": "model.head.dense.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "77d0ed479cad468518e7d5599c9d9e5c39e4a46a40af81e5cef39ce51061820f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "77d0ed479cad468518e7d5599c9d9e5c39e4a46a40af81e5cef39ce51061820f" + }, + "external": true, + "passed": true + }, + { + "name": "model.head.norm.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "c0ff8da291b1f26ee838e7a83ccd3fdf275587e1fccb5c2e45e906266c2bfc7d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768 + ], + "bytes": 3072, + "sha256": "c0ff8da291b1f26ee838e7a83ccd3fdf275587e1fccb5c2e45e906266c2bfc7d" + }, + "external": true, + "passed": true + }, + { + "name": "model.classifier.weight", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 3, + 768 + ], + "bytes": 9216, + "sha256": "3d66e71783c13bfe619ee3e3b8c0dd828d94c1ddf417ee7640b4efb2ff0a213d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 3, + 768 + ], + "bytes": 9216, + "sha256": "3d66e71783c13bfe619ee3e3b8c0dd828d94c1ddf417ee7640b4efb2ff0a213d" + }, + "external": true, + "passed": true + }, + { + "name": "model.classifier.bias", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 3 + ], + "bytes": 12, + "sha256": "f879d7ffe60d77c365e80b914d47bbc79ad78ecefc9637cd48782d20fa166e11" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 3 + ], + "bytes": 12, + "sha256": "f879d7ffe60d77c365e80b914d47bbc79ad78ecefc9637cd48782d20fa166e11" + }, + "external": false, + "passed": true + }, + { + "name": "val_4", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [], + "bytes": 8, + "sha256": "af5570f5a1810b7af78caf4bc70a660f0df51e42baf91d4de5b2328de0e83dfc" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [], + "bytes": 8, + "sha256": "af5570f5a1810b7af78caf4bc70a660f0df51e42baf91d4de5b2328de0e83dfc" + }, + "external": false, + "passed": true + }, + { + "name": "val_6", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [], + "bytes": 8, + "sha256": "7c9fa136d4413fa6173637e883b6998d32e1d675f88cddff9dcbcf331820f4b8" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [], + "bytes": 8, + "sha256": "7c9fa136d4413fa6173637e883b6998d32e1d675f88cddff9dcbcf331820f4b8" + }, + "external": false, + "passed": true + }, + { + "name": "detach", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "e00e5eb9444182f352323374ef4e08ebcb784725fdd4fd612d7730540b3e0c8c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "e00e5eb9444182f352323374ef4e08ebcb784725fdd4fd612d7730540b3e0c8c" + }, + "external": false, + "passed": true + }, + { + "name": "val_24", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "1765f39d5e2ee8b61268a34d74420f2e90396e02ec52875c8cd92409be6ceb67" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "1765f39d5e2ee8b61268a34d74420f2e90396e02ec52875c8cd92409be6ceb67" + }, + "external": true, + "passed": true + }, + { + "name": "expand_1", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1, + 32, + 1 + ], + "bytes": 128, + "sha256": "adb41194debe8e6c185d47754f68cc2dd59960a03cb4ccbc71acd14b860c4604" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1, + 32, + 1 + ], + "bytes": 128, + "sha256": "adb41194debe8e6c185d47754f68cc2dd59960a03cb4ccbc71acd14b860c4604" + }, + "external": false, + "passed": true + }, + { + "name": "val_52", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "af5570f5a1810b7af78caf4bc70a660f0df51e42baf91d4de5b2328de0e83dfc" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "af5570f5a1810b7af78caf4bc70a660f0df51e42baf91d4de5b2328de0e83dfc" + }, + "external": false, + "passed": true + }, + { + "name": "val_56", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "9d4ac218fb54041e3a70a8e14db1ea1af9f570f4842b4702ef9323f1ad8f0ec4" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "9d4ac218fb54041e3a70a8e14db1ea1af9f570f4842b4702ef9323f1ad8f0ec4" + }, + "external": false, + "passed": true + }, + { + "name": "val_60", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "35be322d094f9d154a8aba4733b8497f180353bd7ae7b0a15f90b586b549f28b" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "35be322d094f9d154a8aba4733b8497f180353bd7ae7b0a15f90b586b549f28b" + }, + "external": false, + "passed": true + }, + { + "name": "val_68", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "6a69a6cc7473a16302890cd2a9e93e347281f6ea0e1bb784e589753bed0b3324" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "6a69a6cc7473a16302890cd2a9e93e347281f6ea0e1bb784e589753bed0b3324" + }, + "external": false, + "passed": true + }, + { + "name": "val_116", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1 + ], + "bytes": 4, + "sha256": "948945eecf4a060f48a79e9fadcd501373212c7f5be4fa5ac1ac216f2469c043" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1 + ], + "bytes": 4, + "sha256": "948945eecf4a060f48a79e9fadcd501373212c7f5be4fa5ac1ac216f2469c043" + }, + "external": false, + "passed": true + }, + { + "name": "val_128", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "e3e58e351bc840ce171c8c549f2b157733fda22bf529bc08d8d4984af1f77b8d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "e3e58e351bc840ce171c8c549f2b157733fda22bf529bc08d8d4984af1f77b8d" + }, + "external": true, + "passed": true + }, + { + "name": "val_131", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "8ecb7ab058dc9d0b126a1b4054e56c209308b9b09e6385b2d9ee0666e8d82f71" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "8ecb7ab058dc9d0b126a1b4054e56c209308b9b09e6385b2d9ee0666e8d82f71" + }, + "external": true, + "passed": true + }, + { + "name": "val_140", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "dc259a27849660a9b2c360a89ccf44cbfe24286b93ee606d2fbcbd3781f98a59" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "dc259a27849660a9b2c360a89ccf44cbfe24286b93ee606d2fbcbd3781f98a59" + }, + "external": true, + "passed": true + }, + { + "name": "val_143", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "81a16b7c72612a32c499c95377c6c1599731ea54467ad4e9fee1b9743df6f46e" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "81a16b7c72612a32c499c95377c6c1599731ea54467ad4e9fee1b9743df6f46e" + }, + "external": true, + "passed": true + }, + { + "name": "val_241", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "e5d4f7957244470c49a9567cfa1dd8d4eb9718b9f891426a7056df984871fc4b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "e5d4f7957244470c49a9567cfa1dd8d4eb9718b9f891426a7056df984871fc4b" + }, + "external": true, + "passed": true + }, + { + "name": "val_244", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "71562fb1c132331d9d4262d60755d737f3c28710d2ff44290d49102c1c412d45" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "71562fb1c132331d9d4262d60755d737f3c28710d2ff44290d49102c1c412d45" + }, + "external": true, + "passed": true + }, + { + "name": "val_252", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "097334de817d798b383b662c52a6b2bdbaefd8c8f02d202276372d4512f3a594" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "097334de817d798b383b662c52a6b2bdbaefd8c8f02d202276372d4512f3a594" + }, + "external": true, + "passed": true + }, + { + "name": "val_255", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "4f9ff48c5539f46a3917db2568d8b731ec1a11ae6cfe84238cb18f8649d6c75b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "4f9ff48c5539f46a3917db2568d8b731ec1a11ae6cfe84238cb18f8649d6c75b" + }, + "external": true, + "passed": true + }, + { + "name": "val_353", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "5614412bcca96639e057df96b421b43150036aa25aa7f62f226743e18a481fd3" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "5614412bcca96639e057df96b421b43150036aa25aa7f62f226743e18a481fd3" + }, + "external": true, + "passed": true + }, + { + "name": "val_356", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d58020ffc4ea7d2499c8ab93042b8e3defa50b0de7ea5bdb096dff451be33d9d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d58020ffc4ea7d2499c8ab93042b8e3defa50b0de7ea5bdb096dff451be33d9d" + }, + "external": true, + "passed": true + }, + { + "name": "val_364", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "a7071101955b62b47bc4a477d5b5796230fce630fe445c58b138bb4ce4d889a1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "a7071101955b62b47bc4a477d5b5796230fce630fe445c58b138bb4ce4d889a1" + }, + "external": true, + "passed": true + }, + { + "name": "val_367", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "644bf166e116b96699148c6d99ac3089b0732b446cc8cb69fc87b4b0c3543b4f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "644bf166e116b96699148c6d99ac3089b0732b446cc8cb69fc87b4b0c3543b4f" + }, + "external": true, + "passed": true + }, + { + "name": "val_465", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "d06cb6edbf50417b452e9594d2ef0658340aebdc184b478870f84ad70d1c9e76" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "d06cb6edbf50417b452e9594d2ef0658340aebdc184b478870f84ad70d1c9e76" + }, + "external": true, + "passed": true + }, + { + "name": "val_468", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "14f61658f5afaca9e9d2222520a0e81eeb36fd153802b752ebe8772f94eab363" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "14f61658f5afaca9e9d2222520a0e81eeb36fd153802b752ebe8772f94eab363" + }, + "external": true, + "passed": true + }, + { + "name": "val_476", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "5521aacc4e2b85379a567b2f9fcc07a6da375b47b242cb0b5d9437a388819e1c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "5521aacc4e2b85379a567b2f9fcc07a6da375b47b242cb0b5d9437a388819e1c" + }, + "external": true, + "passed": true + }, + { + "name": "val_479", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "61a9d8c04373c45fd7157c9487562cacd90fe2fa66209c2ab45dad56f16df8eb" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "61a9d8c04373c45fd7157c9487562cacd90fe2fa66209c2ab45dad56f16df8eb" + }, + "external": true, + "passed": true + }, + { + "name": "val_577", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "728e2e8d0f572afd8e4c1d80ba5710635543d488fd7b70c252877210ada22e57" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "728e2e8d0f572afd8e4c1d80ba5710635543d488fd7b70c252877210ada22e57" + }, + "external": true, + "passed": true + }, + { + "name": "val_580", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a6a56a931fb452092bd752c560630871837180777b279cbff1918221edec9ac9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a6a56a931fb452092bd752c560630871837180777b279cbff1918221edec9ac9" + }, + "external": true, + "passed": true + }, + { + "name": "val_588", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "bb766c89059b5d32c1e8192e148a44b1c579dc76c2cc2ff022e222ad557889b3" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "bb766c89059b5d32c1e8192e148a44b1c579dc76c2cc2ff022e222ad557889b3" + }, + "external": true, + "passed": true + }, + { + "name": "val_591", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "83d6d72e2e2fdf9fee484bc29a65e997d39eebdc38f47e054edc3e282978487c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "83d6d72e2e2fdf9fee484bc29a65e997d39eebdc38f47e054edc3e282978487c" + }, + "external": true, + "passed": true + }, + { + "name": "val_689", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "4b5bfdc5053f79d8ab987f80f3b8bb2b9121df7b0f56d1637efed33a5715ef85" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "4b5bfdc5053f79d8ab987f80f3b8bb2b9121df7b0f56d1637efed33a5715ef85" + }, + "external": true, + "passed": true + }, + { + "name": "val_692", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "1c634eb39e05e45421e9355bcabeb61308155628f8f9404b47ea66d22d615ef1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "1c634eb39e05e45421e9355bcabeb61308155628f8f9404b47ea66d22d615ef1" + }, + "external": true, + "passed": true + }, + { + "name": "val_700", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "aa2e0479db6b03016645764695e0cb1f74be90dfd627d9d2713682a789f37ff6" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "aa2e0479db6b03016645764695e0cb1f74be90dfd627d9d2713682a789f37ff6" + }, + "external": true, + "passed": true + }, + { + "name": "val_703", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f519788ceab3b8ad6d48a7b84a91bdd4296ac1e3e4cea0884690f996b8676442" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f519788ceab3b8ad6d48a7b84a91bdd4296ac1e3e4cea0884690f996b8676442" + }, + "external": true, + "passed": true + }, + { + "name": "val_801", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ef6592e754a943c60f47eed2a8cdad5eb0dbf6ffd991a240942cff5fee038840" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ef6592e754a943c60f47eed2a8cdad5eb0dbf6ffd991a240942cff5fee038840" + }, + "external": true, + "passed": true + }, + { + "name": "val_804", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f313208def1bd0fd0de73c1d6c548448f312f3b78c01e160c96d492feee0dfd9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f313208def1bd0fd0de73c1d6c548448f312f3b78c01e160c96d492feee0dfd9" + }, + "external": true, + "passed": true + }, + { + "name": "val_812", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "d6e032c1a457e3b8a3a57b14d326baf0b7578b88d8940dba9792a9b0e10b1e85" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "d6e032c1a457e3b8a3a57b14d326baf0b7578b88d8940dba9792a9b0e10b1e85" + }, + "external": true, + "passed": true + }, + { + "name": "val_815", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "bbe1cd2dfa98a452bfc201211dbe561188f367cccef08a2b776775d14fd0b126" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "bbe1cd2dfa98a452bfc201211dbe561188f367cccef08a2b776775d14fd0b126" + }, + "external": true, + "passed": true + }, + { + "name": "val_913", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ffd8dda8ff197a68f15755f30f30bd17233bd5a10ad62d03d19b4306aab28eb9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ffd8dda8ff197a68f15755f30f30bd17233bd5a10ad62d03d19b4306aab28eb9" + }, + "external": true, + "passed": true + }, + { + "name": "val_916", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "37b0046c876d835567e3a5317a38c0ae7b3c6d45f926f24c6be3df0a73e36d9e" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "37b0046c876d835567e3a5317a38c0ae7b3c6d45f926f24c6be3df0a73e36d9e" + }, + "external": true, + "passed": true + }, + { + "name": "val_924", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "e44fd4eaf5c3e4a278be5e6d55ee9b7cc981ec1b6f82a4a5120678683e77842b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "e44fd4eaf5c3e4a278be5e6d55ee9b7cc981ec1b6f82a4a5120678683e77842b" + }, + "external": true, + "passed": true + }, + { + "name": "val_927", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "4b33d0562f09df57a1e41e14b5ce574cd3e6b759f17ddcfaa27bde870848442c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "4b33d0562f09df57a1e41e14b5ce574cd3e6b759f17ddcfaa27bde870848442c" + }, + "external": true, + "passed": true + }, + { + "name": "val_1025", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "98032d09775f8e6edb065644f1cfefb3fe54d5d12e90f9610ce3b5895d556274" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "98032d09775f8e6edb065644f1cfefb3fe54d5d12e90f9610ce3b5895d556274" + }, + "external": true, + "passed": true + }, + { + "name": "val_1028", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "30e7396c81a728cdbad90bc00a0888f83128635f102c8fa9f26602c3a4773fc4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "30e7396c81a728cdbad90bc00a0888f83128635f102c8fa9f26602c3a4773fc4" + }, + "external": true, + "passed": true + }, + { + "name": "val_1036", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "c15ae30b6f93add6a6044942d5be7b4fd979cb8840b863f38fa30fca1db0e692" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "c15ae30b6f93add6a6044942d5be7b4fd979cb8840b863f38fa30fca1db0e692" + }, + "external": true, + "passed": true + }, + { + "name": "val_1039", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a94a018ace24ea3720d3c1174e69a94dc65604ddb7d3cd88f11f8dc130517720" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a94a018ace24ea3720d3c1174e69a94dc65604ddb7d3cd88f11f8dc130517720" + }, + "external": true, + "passed": true + }, + { + "name": "val_1137", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "36a23c2ca310da3c9c4db524eca4f29a52ad1df11be416c6ea26ae9b3a71f2c5" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "36a23c2ca310da3c9c4db524eca4f29a52ad1df11be416c6ea26ae9b3a71f2c5" + }, + "external": true, + "passed": true + }, + { + "name": "val_1140", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a392f719876e9dc7e4b72714912fe4252193385974a7e2d0cecb5264ee229a19" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "a392f719876e9dc7e4b72714912fe4252193385974a7e2d0cecb5264ee229a19" + }, + "external": true, + "passed": true + }, + { + "name": "val_1148", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "8e196a46834478ffa55709c06ac9578a065de6a72a1d2c99d6f183ba22fefa02" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "8e196a46834478ffa55709c06ac9578a065de6a72a1d2c99d6f183ba22fefa02" + }, + "external": true, + "passed": true + }, + { + "name": "val_1151", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "5432a4590e811498cf009b0c1a5e10290293edc63ffcce75a48e7945dd4cb522" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "5432a4590e811498cf009b0c1a5e10290293edc63ffcce75a48e7945dd4cb522" + }, + "external": true, + "passed": true + }, + { + "name": "val_1249", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "a7ea00b9c61e9e4d5c270e8ecc61dfc98dd38d14c509ac3f40cf4b7c95ee71c4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "a7ea00b9c61e9e4d5c270e8ecc61dfc98dd38d14c509ac3f40cf4b7c95ee71c4" + }, + "external": true, + "passed": true + }, + { + "name": "val_1252", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "80cbf37a7cca157b98dcaf4530adb2bc81102d77b3013df388184ec7a05800cd" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "80cbf37a7cca157b98dcaf4530adb2bc81102d77b3013df388184ec7a05800cd" + }, + "external": true, + "passed": true + }, + { + "name": "val_1260", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "78d077323bb28eaaad50b97790617dd3003091be44efd6c12505f321f470993f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "78d077323bb28eaaad50b97790617dd3003091be44efd6c12505f321f470993f" + }, + "external": true, + "passed": true + }, + { + "name": "val_1263", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "ebee9cd09721be07e63f909754467e0d2e8496c85728cd2042f8ef67d8c41c7d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "ebee9cd09721be07e63f909754467e0d2e8496c85728cd2042f8ef67d8c41c7d" + }, + "external": true, + "passed": true + }, + { + "name": "val_1361", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ce7fa36ccc306ae8d65ec6b1409fddc662442208f7440b3642f12151fc4e4a80" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ce7fa36ccc306ae8d65ec6b1409fddc662442208f7440b3642f12151fc4e4a80" + }, + "external": true, + "passed": true + }, + { + "name": "val_1364", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "7846a8cb9e40106940a0d828ccb9b9115728f4a383ce508f0e46ad798f117a7f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "7846a8cb9e40106940a0d828ccb9b9115728f4a383ce508f0e46ad798f117a7f" + }, + "external": true, + "passed": true + }, + { + "name": "val_1372", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "7989c676498349efd3aeb496dbcc0c4b253724bd72d2cf343109b0eb24140e22" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "7989c676498349efd3aeb496dbcc0c4b253724bd72d2cf343109b0eb24140e22" + }, + "external": true, + "passed": true + }, + { + "name": "val_1375", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "c71b276fa63e7e00adeaf14b0a59c0e04305b774b497ed21d87faee4131268db" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "c71b276fa63e7e00adeaf14b0a59c0e04305b774b497ed21d87faee4131268db" + }, + "external": true, + "passed": true + }, + { + "name": "val_1473", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ee3cb96d693f39268f52e016ca6a5186902876f0c52c1b27c53f0c40cacdf74d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "ee3cb96d693f39268f52e016ca6a5186902876f0c52c1b27c53f0c40cacdf74d" + }, + "external": true, + "passed": true + }, + { + "name": "val_1476", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "bc1ef53ca3bc9f8dcac319da237ca65c8ae2f254577468017c46a01a39b565ac" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "bc1ef53ca3bc9f8dcac319da237ca65c8ae2f254577468017c46a01a39b565ac" + }, + "external": true, + "passed": true + }, + { + "name": "val_1484", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "c834c09b25e38f5b2832212a02933012bbd1d94c2de03633b79c4e35a9995a54" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "c834c09b25e38f5b2832212a02933012bbd1d94c2de03633b79c4e35a9995a54" + }, + "external": true, + "passed": true + }, + { + "name": "val_1487", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "be1184abcf55828ca61de038a993890158404f3f0b99327b096bdb767f0487c1" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "be1184abcf55828ca61de038a993890158404f3f0b99327b096bdb767f0487c1" + }, + "external": true, + "passed": true + }, + { + "name": "val_1585", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "236fa7b232e5a184540bcfa388a2a10e321d3fd8d38c57c029ff6b4df545539c" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "236fa7b232e5a184540bcfa388a2a10e321d3fd8d38c57c029ff6b4df545539c" + }, + "external": true, + "passed": true + }, + { + "name": "val_1588", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "2c9d6a92b8835f9b96bdc7163603d0f4163dd75f40a8267260b036defb53df24" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "2c9d6a92b8835f9b96bdc7163603d0f4163dd75f40a8267260b036defb53df24" + }, + "external": true, + "passed": true + }, + { + "name": "val_1596", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "b9f4f69481af0c60a3b211f17185559db75c88df9ae2f02df2c5d5167974509d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "b9f4f69481af0c60a3b211f17185559db75c88df9ae2f02df2c5d5167974509d" + }, + "external": true, + "passed": true + }, + { + "name": "val_1599", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d7cfd772f3ab5cf07cd9dc919ca81a514d3cc8ca8c4087feb95f398cf64c6cf9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d7cfd772f3ab5cf07cd9dc919ca81a514d3cc8ca8c4087feb95f398cf64c6cf9" + }, + "external": true, + "passed": true + }, + { + "name": "val_1697", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "f20d3dc40e93f093834765294b75a8d15660bfc1959d531818a616291b7b8442" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "f20d3dc40e93f093834765294b75a8d15660bfc1959d531818a616291b7b8442" + }, + "external": true, + "passed": true + }, + { + "name": "val_1700", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d25d70fc6901e433fe5492524791467f79fcbf6a11deef299197cf375a17efd5" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d25d70fc6901e433fe5492524791467f79fcbf6a11deef299197cf375a17efd5" + }, + "external": true, + "passed": true + }, + { + "name": "val_1708", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "ba354e24a72a8a23e7f58361ccf510caaf0d9b7d9923e39580a3a08ba7d80da3" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "ba354e24a72a8a23e7f58361ccf510caaf0d9b7d9923e39580a3a08ba7d80da3" + }, + "external": true, + "passed": true + }, + { + "name": "val_1711", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fed35fdf6c4c5ceff51caee65a57fc3b5e904bdf0b0ccd2b44a8d1eeeadd2b9d" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fed35fdf6c4c5ceff51caee65a57fc3b5e904bdf0b0ccd2b44a8d1eeeadd2b9d" + }, + "external": true, + "passed": true + }, + { + "name": "val_1809", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "01e078ed1dd0e0d3dacca2f2ed141f53e56b5351b36f464ee46df54359add709" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "01e078ed1dd0e0d3dacca2f2ed141f53e56b5351b36f464ee46df54359add709" + }, + "external": true, + "passed": true + }, + { + "name": "val_1812", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d33c7aca0b242cc8c20b582fc15e2be4677ecb06ffd13a2bfd2ca8f2b326d3f7" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "d33c7aca0b242cc8c20b582fc15e2be4677ecb06ffd13a2bfd2ca8f2b326d3f7" + }, + "external": true, + "passed": true + }, + { + "name": "val_1820", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "96e833cda7bd5e4e07d8b2f683bd320fbab937b28d5ef9c40de3e6bf0552bed4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "96e833cda7bd5e4e07d8b2f683bd320fbab937b28d5ef9c40de3e6bf0552bed4" + }, + "external": true, + "passed": true + }, + { + "name": "val_1823", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "0de0b167335f6b43428187fd245a46af30715e07e9cb9967a5e79285801589ee" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "0de0b167335f6b43428187fd245a46af30715e07e9cb9967a5e79285801589ee" + }, + "external": true, + "passed": true + }, + { + "name": "val_1921", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "c60c45ecf572a04b2b8d1f46f6a8cc43e06b9408df491af96f0cf417dfb9250b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "c60c45ecf572a04b2b8d1f46f6a8cc43e06b9408df491af96f0cf417dfb9250b" + }, + "external": true, + "passed": true + }, + { + "name": "val_1924", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fc62dbdc9941f453697c34f03c58b5aa456bbc4d089d017d33cde4160ccc1227" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fc62dbdc9941f453697c34f03c58b5aa456bbc4d089d017d33cde4160ccc1227" + }, + "external": true, + "passed": true + }, + { + "name": "val_1932", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "36c0124c31b25a28c8513e6fb6e76da4dcd5363608488132dfb6dcd9b32d1089" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "36c0124c31b25a28c8513e6fb6e76da4dcd5363608488132dfb6dcd9b32d1089" + }, + "external": true, + "passed": true + }, + { + "name": "val_1935", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "ada22463330a4c1f4320e35658c172b33991adfeea51364312463422e4e222f5" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "ada22463330a4c1f4320e35658c172b33991adfeea51364312463422e4e222f5" + }, + "external": true, + "passed": true + }, + { + "name": "val_2033", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "70cc5eb6789f65048feabbef791578ba8999378486ac1810a009fdb32011f0ef" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "70cc5eb6789f65048feabbef791578ba8999378486ac1810a009fdb32011f0ef" + }, + "external": true, + "passed": true + }, + { + "name": "val_2036", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "008b89f21bd44be6c8811207df9a06e4a1ca4f31a315a9ea205e2306e311d259" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "008b89f21bd44be6c8811207df9a06e4a1ca4f31a315a9ea205e2306e311d259" + }, + "external": true, + "passed": true + }, + { + "name": "val_2044", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "cce331508f995bb7d29459b9aa1e0d7cfc3d4b33a1c9dff94babb528b8067e65" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "cce331508f995bb7d29459b9aa1e0d7cfc3d4b33a1c9dff94babb528b8067e65" + }, + "external": true, + "passed": true + }, + { + "name": "val_2047", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fc31ba338408e8730eb0b26c63b95e6d1be5e5aeed8b76222e8e446bcb1910a9" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "fc31ba338408e8730eb0b26c63b95e6d1be5e5aeed8b76222e8e446bcb1910a9" + }, + "external": true, + "passed": true + }, + { + "name": "val_2145", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "edf5bb47f0e59e734dc7f987681f3bee65d5ef565b83e16b0141ae469bde4e19" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "edf5bb47f0e59e734dc7f987681f3bee65d5ef565b83e16b0141ae469bde4e19" + }, + "external": true, + "passed": true + }, + { + "name": "val_2148", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "82bc5667cd87d00e1ef2a50bc426c81b349d9101875e71cc9ddbc71a2c6c6b8b" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "82bc5667cd87d00e1ef2a50bc426c81b349d9101875e71cc9ddbc71a2c6c6b8b" + }, + "external": true, + "passed": true + }, + { + "name": "val_2156", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "da38d2976197d7cc0b4254db950bf05872154a7348ac342be82a4cf5e73f7ffc" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "da38d2976197d7cc0b4254db950bf05872154a7348ac342be82a4cf5e73f7ffc" + }, + "external": true, + "passed": true + }, + { + "name": "val_2159", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f8bfd5a29a791953706db4054a6d919afad6c60ebcdb895073e7a07b9114a3e8" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "f8bfd5a29a791953706db4054a6d919afad6c60ebcdb895073e7a07b9114a3e8" + }, + "external": true, + "passed": true + }, + { + "name": "val_2257", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "2a3b85b4d57775aa86a0f5c002bbd8618578da4d936fd3f2f9a8c45fe9e8bbef" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "2a3b85b4d57775aa86a0f5c002bbd8618578da4d936fd3f2f9a8c45fe9e8bbef" + }, + "external": true, + "passed": true + }, + { + "name": "val_2260", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "eb7e563deb048799ecf197c085e056c7119fb29be385b8ddfc738403e629c182" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "eb7e563deb048799ecf197c085e056c7119fb29be385b8ddfc738403e629c182" + }, + "external": true, + "passed": true + }, + { + "name": "val_2268", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "371251338421ded1360e2d94b3f7889cad6532cacf1ca857050c319ff98e2615" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "371251338421ded1360e2d94b3f7889cad6532cacf1ca857050c319ff98e2615" + }, + "external": true, + "passed": true + }, + { + "name": "val_2271", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "34bc0a55945385aebcdb22c14caa718ce61ba0abad31962b5d4fce4cad7849f4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "34bc0a55945385aebcdb22c14caa718ce61ba0abad31962b5d4fce4cad7849f4" + }, + "external": true, + "passed": true + }, + { + "name": "val_2369", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "47304e7e777c0e77c355ecd16b24c0f902653c4eb7327d699b2b96c760d2a1b4" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "47304e7e777c0e77c355ecd16b24c0f902653c4eb7327d699b2b96c760d2a1b4" + }, + "external": true, + "passed": true + }, + { + "name": "val_2372", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "21a15f49ade90cf9991f8eb57e6cc21ad04bbb1fbe030cae4a4b5453a9f9b3b6" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "21a15f49ade90cf9991f8eb57e6cc21ad04bbb1fbe030cae4a4b5453a9f9b3b6" + }, + "external": true, + "passed": true + }, + { + "name": "val_2380", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "76379add6ddbaee76a60f5af868e43c58b79fda94b08321923f7fe3c980db2e2" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "76379add6ddbaee76a60f5af868e43c58b79fda94b08321923f7fe3c980db2e2" + }, + "external": true, + "passed": true + }, + { + "name": "val_2383", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "edd7b1b1fcb1313c4553cefc7c3c0fd41bfc088c90b91b8c0120635344069bff" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "edd7b1b1fcb1313c4553cefc7c3c0fd41bfc088c90b91b8c0120635344069bff" + }, + "external": true, + "passed": true + }, + { + "name": "val_2481", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "1206c826a833c8bc5065c67c951ffded36e78b5ab4542d5b361e56ee57aa429f" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 768 + ], + "bytes": 2359296, + "sha256": "1206c826a833c8bc5065c67c951ffded36e78b5ab4542d5b361e56ee57aa429f" + }, + "external": true, + "passed": true + }, + { + "name": "val_2484", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "475c5ef0d11462f712f17503b9ffd115ab59c506eb7f2830a0e8f548d952a679" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 768, + 2304 + ], + "bytes": 7077888, + "sha256": "475c5ef0d11462f712f17503b9ffd115ab59c506eb7f2830a0e8f548d952a679" + }, + "external": true, + "passed": true + }, + { + "name": "val_2492", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "7798fc32a103d4939230fb708376ace81655ac476b8c748d59eede47113e2e98" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [ + 1152, + 768 + ], + "bytes": 3538944, + "sha256": "7798fc32a103d4939230fb708376ace81655ac476b8c748d59eede47113e2e98" + }, + "external": true, + "passed": true + }, + { + "name": "val_8", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "7c9fa136d4413fa6173637e883b6998d32e1d675f88cddff9dcbcf331820f4b8" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "7c9fa136d4413fa6173637e883b6998d32e1d675f88cddff9dcbcf331820f4b8" + }, + "external": false, + "passed": true + }, + { + "name": "val_9", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "d86e8112f3c4c4442126f8e9f44f16867da487f29052bf91b810457db34209a4" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "d86e8112f3c4c4442126f8e9f44f16867da487f29052bf91b810457db34209a4" + }, + "external": false, + "passed": true + }, + { + "name": "val_27", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "12a3ae445661ce5dee78d0650d33362dec29c4f82af05e7e57fb595bbbacf0ca" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "12a3ae445661ce5dee78d0650d33362dec29c4f82af05e7e57fb595bbbacf0ca" + }, + "external": false, + "passed": true + }, + { + "name": "val_29", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "220dde27afee4d537cac96d85d6546f825153b90e828931b74e103807541bc42" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "220dde27afee4d537cac96d85d6546f825153b90e828931b74e103807541bc42" + }, + "external": false, + "passed": true + }, + { + "name": "val_30", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "a06f129fc52abf6085679d7cd71dc41ec7580c7f5f73efef6d02dde22bb00994" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "a06f129fc52abf6085679d7cd71dc41ec7580c7f5f73efef6d02dde22bb00994" + }, + "external": false, + "passed": true + }, + { + "name": "val_106", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "81b37f7f8b78762feb067131152a700548754f937a0668c1f327d2ee1a2d2801" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "81b37f7f8b78762feb067131152a700548754f937a0668c1f327d2ee1a2d2801" + }, + "external": false, + "passed": true + }, + { + "name": "val_108", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "e6ad6c9a3a3b7658c35bacf6553fcb8ffe34387534a648fe18f875b8f7a86ddb" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "e6ad6c9a3a3b7658c35bacf6553fcb8ffe34387534a648fe18f875b8f7a86ddb" + }, + "external": false, + "passed": true + }, + { + "name": "val_126", + "original": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "8285d24b1ca2820ef8beda7b35738ebb281870d192a06328d284dc3177f4fe31" + }, + "final": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 1 + ], + "bytes": 8, + "sha256": "8285d24b1ca2820ef8beda7b35738ebb281870d192a06328d284dc3177f4fe31" + }, + "external": false, + "passed": true + }, + { + "name": "val_133", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "9c16095ed50248c0e1658be07f5bafe5dc11382843a99c6fb3a87c141808ceeb" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "9c16095ed50248c0e1658be07f5bafe5dc11382843a99c6fb3a87c141808ceeb" + }, + "external": false, + "passed": true + }, + { + "name": "val_138", + "original": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "d99e58435243d9fef9c88273b8d553b4fba4d0baf8009d29eae74fa99e0d9f57" + }, + "final": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "d99e58435243d9fef9c88273b8d553b4fba4d0baf8009d29eae74fa99e0d9f57" + }, + "external": false, + "passed": true + } + ], + "removed_original_initializers": [ + { + "name": "val_17", + "identity": { + "dtype": "FLOAT", + "dtype_id": 1, + "shape": [], + "bytes": 4, + "sha256": "abc4c4e7eee2ce04c6be04358adf5b3d0958255a80f712a8b6d4f4fdda588276" + } + }, + { + "name": "val_20", + "identity": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [], + "bytes": 8, + "sha256": "a06f129fc52abf6085679d7cd71dc41ec7580c7f5f73efef6d02dde22bb00994" + } + }, + { + "name": "val_2506", + "identity": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 2 + ], + "bytes": 16, + "sha256": "0c730b69905c5ef7a4ca5269f72365400bde2dd2c04eaf9bbb3d1c4a265a0131" + } + }, + { + "name": "val_2507", + "identity": { + "dtype": "INT64", + "dtype_id": 7, + "shape": [ + 2 + ], + "bytes": 16, + "sha256": "9d34149fbd1fe777eb238799054c8cbfbce372255f219f8740838def9bfd02db" + } + } + ] +} diff --git a/evaluation/engines/vela-modality-blocked-b1-native-gpu-32k-v1.json b/evaluation/engines/vela-modality-blocked-b1-native-gpu-32k-v1.json new file mode 100644 index 0000000000000000000000000000000000000000..4c3b73591a3337ee61582f4238652ed18d5d6686 --- /dev/null +++ b/evaluation/engines/vela-modality-blocked-b1-native-gpu-32k-v1.json @@ -0,0 +1,95 @@ +{ + "engine": "onnx", + "gpu": true, + "reference_sha256": "2942ef90b1dd97f90e97793e663d4a8f993ef66922b43c818261121fd5f68cd4", + "weights_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "parity": [ + { + "tokens": 9, + "max_abs_score_error": 4.172325134277344e-07, + "argmax_same": true, + "native_ms": 392.0625450555235, + "input_sha256": "28a5f5d83deb4fffe1ba2c2ddd3e7d8cb5d09c2aa96080454ae3dade0c99406a" + }, + { + "tokens": 10, + "max_abs_score_error": 2.3283064365386963e-08, + "argmax_same": true, + "native_ms": 15.951868961565197, + "input_sha256": "c2f6fc42badff8e41964d9d96ab2ec7ca95386bb1f9de6498e9e07e3ee40a743" + }, + { + "tokens": 11, + "max_abs_score_error": 2.7284841053187847e-11, + "argmax_same": true, + "native_ms": 15.637201024219394, + "input_sha256": "d1151ac839a0385c1658a5be98b53a796c7e040c22eaeb8310e11a118132b10a" + }, + { + "tokens": 10, + "max_abs_score_error": 3.055333763768431e-13, + "argmax_same": true, + "native_ms": 14.953606063500047, + "input_sha256": "df7c0764375e479f20c7b237daa80a3526c0d74afe23d8dbc5d678b2f368b211" + }, + { + "tokens": 63, + "max_abs_score_error": 7.130438461899757e-10, + "argmax_same": true, + "native_ms": 15.710555016994476, + "input_sha256": "ae2ff5297ca5473f882a421ea7055e249c6096ce24e40a929ade9dc1b0760862" + }, + { + "tokens": 64, + "max_abs_score_error": 1.6007106751203537e-10, + "argmax_same": true, + "native_ms": 15.440835035406053, + "input_sha256": "2ee0d8dc7879369d6e6827e3be54f03e794dfcee1670afec37c4957e983f0b87" + }, + { + "tokens": 65, + "max_abs_score_error": 2.954038791358471e-09, + "argmax_same": true, + "native_ms": 15.359019045718014, + "input_sha256": "d0e67fca708f0264df1513f7b0b1508692a6c42eb7bf35a662441e8bf7119d5c" + }, + { + "tokens": 512, + "max_abs_score_error": 3.956301952712238e-11, + "argmax_same": true, + "native_ms": 25.833902996964753, + "input_sha256": "b1e6aa41d7c4c3d327fff16c873fdb165552ed315254e0c7756175bfe0e1984b" + }, + { + "tokens": 4096, + "max_abs_score_error": 1.1920928955078125e-07, + "argmax_same": true, + "native_ms": 196.1183260427788, + "input_sha256": "38f565a661a12dc4af985b73b73b86f4e0731f74334a9fde7d9b7e5f5b0d9394" + }, + { + "tokens": 8192, + "max_abs_score_error": 5.122274160385132e-09, + "argmax_same": true, + "native_ms": 779.2861249763519, + "input_sha256": "ec217f8d64df4984535686f985036008794c15282a464fe0cba1092825b2c788" + }, + { + "tokens": 16384, + "max_abs_score_error": 1.0986695997416973e-09, + "argmax_same": true, + "native_ms": 2589.4851229386404, + "input_sha256": "2aa1c56673d987a494f1600f6cd36baf4f4e6b7bccf1d2892d0df9e50a3c5f69" + }, + { + "tokens": 32768, + "max_abs_score_error": 1.2945383787155151e-07, + "argmax_same": true, + "native_ms": 9555.754297995009, + "input_sha256": "fb5e1bdb22c71316ad3c2a458e5a7dcf1453babbf643d2c0e927af53f1c08c54" + } + ], + "overflow_rejected": true, + "reopen": true, + "label_mismatch_rejected": true +} diff --git a/evaluation/engines/vela-modality-blocked-b1-onnx-gpu-v1.json b/evaluation/engines/vela-modality-blocked-b1-onnx-gpu-v1.json new file mode 100644 index 0000000000000000000000000000000000000000..0c335a6c6a129647cf1d75766107cb86ec81e365 --- /dev/null +++ b/evaluation/engines/vela-modality-blocked-b1-onnx-gpu-v1.json @@ -0,0 +1,143 @@ +{ + "qualified": true, + "weights_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "reference_sha256": "2942ef90b1dd97f90e97793e663d4a8f993ef66922b43c818261121fd5f68cd4", + "graph_sha256": "74a08c52ce47f522071fd71acee080c0a3a74ad12e3d5fb19ae1dfd244f1059c", + "precision": "FP32 weights and computation, exact query-block attention", + "probability_atol": 0.0002, + "probability_rtol": 0.0001, + "exact_BIO_required": false, + "exact_argmax_required": true, + "task": "sequence-classification", + "runtime": "1.22.1", + "rows": [ + { + "case": "b1s2", + "valid_tokens": [ + 2 + ], + "max_logit_error": 5.4836273193359375e-06, + "max_probability_error": 3.259629011154175e-09, + "argmax_flips": 0, + "first_inference_ms": 416.892733075656, + "probability_gate": true + }, + { + "case": "b1s63", + "valid_tokens": [ + 63 + ], + "max_logit_error": 2.1696090698242188e-05, + "max_probability_error": 4.6629367034256575e-14, + "argmax_flips": 0, + "first_inference_ms": 32.77761000208557, + "probability_gate": true + }, + { + "case": "b1s64", + "valid_tokens": [ + 64 + ], + "max_logit_error": 1.0967254638671875e-05, + "max_probability_error": 2.0872192862952943e-13, + "argmax_flips": 0, + "first_inference_ms": 33.219670061953366, + "probability_gate": true + }, + { + "case": "b1s65", + "valid_tokens": [ + 65 + ], + "max_logit_error": 1.3589859008789062e-05, + "max_probability_error": 3.623767952376511e-13, + "argmax_flips": 0, + "first_inference_ms": 34.18030601460487, + "probability_gate": true + }, + { + "case": "b1s127", + "valid_tokens": [ + 127 + ], + "max_logit_error": 2.193450927734375e-05, + "max_probability_error": 4.6838977141305804e-11, + "argmax_flips": 0, + "first_inference_ms": 36.99377097655088, + "probability_gate": true + }, + { + "case": "b1s128", + "valid_tokens": [ + 128 + ], + "max_logit_error": 2.288818359375e-05, + "max_probability_error": 2.8421709430404007e-11, + "argmax_flips": 0, + "first_inference_ms": 34.10984796937555, + "probability_gate": true + }, + { + "case": "b1s129", + "valid_tokens": [ + 129 + ], + "max_logit_error": 2.288818359375e-05, + "max_probability_error": 2.6716406864579767e-11, + "argmax_flips": 0, + "first_inference_ms": 34.23338197171688, + "probability_gate": true + }, + { + "case": "b1s512", + "valid_tokens": [ + 512 + ], + "max_logit_error": 5.900859832763672e-06, + "max_probability_error": 5.005858838558197e-09, + "argmax_flips": 0, + "first_inference_ms": 51.06296797748655, + "probability_gate": true + }, + { + "case": "b1s4096", + "valid_tokens": [ + 4096 + ], + "max_logit_error": 2.6702880859375e-05, + "max_probability_error": 7.748603820800781e-07, + "argmax_flips": 0, + "first_inference_ms": 221.17824701126665, + "probability_gate": true + }, + { + "case": "b1s32768", + "valid_tokens": [ + 32768 + ], + "max_logit_error": 2.0325183868408203e-05, + "max_probability_error": 1.1920928955078125e-06, + "argmax_flips": 0, + "first_inference_ms": 9738.925967016257, + "probability_gate": true + } + ], + "diagnostic": false, + "static_batch_size": 1, + "scope": "numerical qualification, not independent quality or isolated latency benchmark", + "providers": [ + "ROCMExecutionProvider", + "CPUExecutionProvider" + ], + "unqualified_batch_rejected": true, + "profile_sha256": "705d6f134ccb875723535ca48d7458f00b53ebd8f7c04452a2c21d7436ec4637", + "profile_providers": { + "ROCMExecutionProvider": 98742, + "CPUExecutionProvider": 12150 + }, + "attention_execution": { + "ROCMExecutionProvider:MatMul": 6732, + "ROCMExecutionProvider:Softmax": 3366 + }, + "attention_on_gpu": true +} diff --git a/evaluation/engines/vela-modality-candle-cpu-v1.json b/evaluation/engines/vela-modality-candle-cpu-v1.json new file mode 100644 index 0000000000000000000000000000000000000000..a4487c4c0244ce1f59498ddb1cd54f707e8d71b4 --- /dev/null +++ b/evaluation/engines/vela-modality-candle-cpu-v1.json @@ -0,0 +1,67 @@ +{ + "engine": "candle", + "gpu": false, + "reference_sha256": "2942ef90b1dd97f90e97793e663d4a8f993ef66922b43c818261121fd5f68cd4", + "weights_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "parity": [ + { + "tokens": 9, + "max_abs_score_error": 1.329183578491211e-05, + "argmax_same": true, + "native_ms": 169.12583902012557, + "input_sha256": "28a5f5d83deb4fffe1ba2c2ddd3e7d8cb5d09c2aa96080454ae3dade0c99406a" + }, + { + "tokens": 10, + "max_abs_score_error": 3.911089152097702e-06, + "argmax_same": true, + "native_ms": 117.81790107488632, + "input_sha256": "c2f6fc42badff8e41964d9d96ab2ec7ca95386bb1f9de6498e9e07e3ee40a743" + }, + { + "tokens": 11, + "max_abs_score_error": 1.5072146197780967e-08, + "argmax_same": true, + "native_ms": 119.22943498939276, + "input_sha256": "d1151ac839a0385c1658a5be98b53a796c7e040c22eaeb8310e11a118132b10a" + }, + { + "tokens": 10, + "max_abs_score_error": 4.339995030022692e-11, + "argmax_same": true, + "native_ms": 119.12891990505159, + "input_sha256": "df7c0764375e479f20c7b237daa80a3526c0d74afe23d8dbc5d678b2f368b211" + }, + { + "tokens": 63, + "max_abs_score_error": 1.1920928955078125e-07, + "argmax_same": true, + "native_ms": 195.58919803239405, + "input_sha256": "ae2ff5297ca5473f882a421ea7055e249c6096ce24e40a929ade9dc1b0760862" + }, + { + "tokens": 64, + "max_abs_score_error": 1.1920928955078125e-07, + "argmax_same": true, + "native_ms": 203.6561779677868, + "input_sha256": "2ee0d8dc7879369d6e6827e3be54f03e794dfcee1670afec37c4957e983f0b87" + }, + { + "tokens": 65, + "max_abs_score_error": 1.2147938832640648e-07, + "argmax_same": true, + "native_ms": 198.97981407120824, + "input_sha256": "d0e67fca708f0264df1513f7b0b1508692a6c42eb7bf35a662441e8bf7119d5c" + }, + { + "tokens": 512, + "max_abs_score_error": 1.90857463167049e-09, + "argmax_same": true, + "native_ms": 1900.0832010060549, + "input_sha256": "b1e6aa41d7c4c3d327fff16c873fdb165552ed315254e0c7756175bfe0e1984b" + } + ], + "overflow_rejected": true, + "reopen": true, + "label_mismatch_rejected": true +} diff --git a/evaluation/engines/vela-modality-standalone-engine-reproduction-v1.json b/evaluation/engines/vela-modality-standalone-engine-reproduction-v1.json new file mode 100644 index 0000000000000000000000000000000000000000..2c4f27da3769c602883dbbc5cd12d92b7bc82bbb --- /dev/null +++ b/evaluation/engines/vela-modality-standalone-engine-reproduction-v1.json @@ -0,0 +1,55 @@ +{ + "qualified": true, + "weights_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "reference_sha256": "bc7fa82ad8bacdb45bf605e5249abc2dca61f081c4d21533c61ff0f535f6e6fa", + "graph_sha256": "74a08c52ce47f522071fd71acee080c0a3a74ad12e3d5fb19ae1dfd244f1059c", + "precision": "FP32 weights and computation, exact query-block attention", + "probability_atol": 0.0002, + "probability_rtol": 0.0001, + "exact_BIO_required": false, + "exact_argmax_required": true, + "task": "sequence-classification", + "runtime": "1.22.1", + "rows": [ + { + "case": "b1s2", + "valid_tokens": [ + 2 + ], + "max_logit_error": 1.1444091796875e-05, + "max_probability_error": 6.100162863731384e-08, + "argmax_flips": 0, + "first_inference_ms": 408.44920300878584, + "probability_gate": true + }, + { + "case": "b1s63", + "valid_tokens": [ + 63 + ], + "max_logit_error": 4.9591064453125e-05, + "max_probability_error": 2.0383694732117874e-13, + "argmax_flips": 0, + "first_inference_ms": 34.35661701951176, + "probability_gate": true + } + ], + "diagnostic": false, + "static_batch_size": 1, + "scope": "numerical qualification, not independent quality or isolated latency benchmark", + "providers": [ + "ROCMExecutionProvider", + "CPUExecutionProvider" + ], + "unqualified_batch_rejected": true, + "profile_sha256": "4d06390d82e383c8dedd92af2c8141dce06b10235fd6480ee7175477e727e246", + "profile_providers": { + "ROCMExecutionProvider": 4190, + "CPUExecutionProvider": 2430 + }, + "attention_execution": { + "ROCMExecutionProvider:MatMul": 88, + "ROCMExecutionProvider:Softmax": 44 + }, + "attention_on_gpu": true +} diff --git a/evaluation/final-baseline-fp32.json b/evaluation/final-baseline-fp32.json new file mode 100644 index 0000000000000000000000000000000000000000..39867e258faf075da2e4cba0e5206eceab9005e7 --- /dev/null +++ b/evaluation/final-baseline-fp32.json @@ -0,0 +1,1445 @@ +{ + "rows": 1411, + "accuracy": 0.4599574769666903, + "macro_f1": 0.3669769557178275, + "macro_f1_present_labels": 0.3669769557178275, + "per_label": { + "AR": { + "support": 643, + "tp": 294, + "fp": 70, + "fn": 349, + "precision": 0.8076923076923077, + "recall": 0.4572317262830482, + "f1": 0.5839126117179742 + }, + "DIFFUSION": { + "support": 384, + "tp": 351, + "fp": 677, + "fn": 33, + "precision": 0.3414396887159533, + "recall": 0.9140625, + "f1": 0.4971671388101983 + }, + "BOTH": { + "support": 384, + "tp": 4, + "fp": 15, + "fn": 380, + "precision": 0.21052631578947367, + "recall": 0.010416666666666666, + "f1": 0.019851116625310174 + } + }, + "confusion": [ + [ + 294, + 335, + 14 + ], + [ + 32, + 351, + 1 + ], + [ + 38, + 342, + 4 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 240, + "accuracy": 0.9666666666666667, + "macro_f1": 0.3276836158192091, + "macro_f1_present_labels": 0.9830508474576272, + "per_label": { + "AR": { + "support": 240, + "tp": 232, + "fp": 0, + "fn": 8, + "precision": 1.0, + "recall": 0.9666666666666667, + "f1": 0.9830508474576272 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 8, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 232, + 0, + 8 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 24, + "fn": 12, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 24, + "tp": 12, + "fp": 24, + "fn": 12, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 24, + "tp": 0, + "fp": 0, + "fn": 24, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 12, + 12, + 0 + ], + [ + 12, + 12, + 0 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 1080, + "accuracy": 0.34629629629629627, + "macro_f1": 0.2216396874882791, + "macro_f1_present_labels": 0.2216396874882791, + "per_label": { + "AR": { + "support": 360, + "tp": 31, + "fp": 46, + "fn": 329, + "precision": 0.4025974025974026, + "recall": 0.08611111111111111, + "f1": 0.14187643020594964 + }, + "DIFFUSION": { + "support": 360, + "tp": 339, + "fp": 653, + "fn": 21, + "precision": 0.34173387096774194, + "recall": 0.9416666666666667, + "f1": 0.5014792899408284 + }, + "BOTH": { + "support": 360, + "tp": 4, + "fp": 7, + "fn": 356, + "precision": 0.36363636363636365, + "recall": 0.011111111111111112, + "f1": 0.0215633423180593 + } + }, + "confusion": [ + [ + 31, + 323, + 6 + ], + [ + 20, + 339, + 1 + ], + [ + 26, + 330, + 4 + ] + ] + }, + "aya-original-human": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 220, + "accuracy": 0.5, + "macro_f1": 0.38727061556329856, + "macro_f1_present_labels": 0.38727061556329856, + "per_label": { + "AR": { + "support": 100, + "tp": 53, + "fp": 22, + "fn": 47, + "precision": 0.7066666666666667, + "recall": 0.53, + "f1": 0.6057142857142858 + }, + "DIFFUSION": { + "support": 60, + "tp": 57, + "fp": 88, + "fn": 3, + "precision": 0.3931034482758621, + "recall": 0.95, + "f1": 0.5560975609756098 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 53, + 47, + 0 + ], + [ + 3, + 57, + 0 + ], + [ + 19, + 41, + 0 + ] + ] + }, + "deu": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 256, + "accuracy": 0.43359375, + "macro_f1": 0.33885209713024284, + "macro_f1_present_labels": 0.33885209713024284, + "per_label": { + "AR": { + "support": 112, + "tp": 39, + "fp": 0, + "fn": 73, + "precision": 1.0, + "recall": 0.3482142857142857, + "f1": 0.5165562913907285 + }, + "DIFFUSION": { + "support": 72, + "tp": 72, + "fp": 144, + "fn": 0, + "precision": 0.3333333333333333, + "recall": 1.0, + "f1": 0.5 + }, + "BOTH": { + "support": 72, + "tp": 0, + "fp": 1, + "fn": 72, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 39, + 72, + 1 + ], + [ + 0, + 72, + 0 + ], + [ + 0, + 72, + 0 + ] + ] + }, + "eng": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 220, + "accuracy": 0.4727272727272727, + "macro_f1": 0.37387387387387383, + "macro_f1_present_labels": 0.37387387387387383, + "per_label": { + "AR": { + "support": 100, + "tp": 46, + "fp": 2, + "fn": 54, + "precision": 0.9583333333333334, + "recall": 0.46, + "f1": 0.6216216216216216 + }, + "DIFFUSION": { + "support": 60, + "tp": 58, + "fp": 114, + "fn": 2, + "precision": 0.3372093023255814, + "recall": 0.9666666666666667, + "f1": 0.5 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 46, + 54, + 0 + ], + [ + 2, + 58, + 0 + ], + [ + 0, + 60, + 0 + ] + ] + }, + "fr": { + "rows": 220, + "accuracy": 0.45454545454545453, + "macro_f1": 0.3553057799633142, + "macro_f1_present_labels": 0.3553057799633142, + "per_label": { + "AR": { + "support": 100, + "tp": 41, + "fp": 5, + "fn": 59, + "precision": 0.8913043478260869, + "recall": 0.41, + "f1": 0.5616438356164384 + }, + "DIFFUSION": { + "support": 60, + "tp": 59, + "fp": 115, + "fn": 1, + "precision": 0.3390804597701149, + "recall": 0.9833333333333333, + "f1": 0.5042735042735043 + }, + "BOTH": { + "support": 60, + "tp": 0, + "fp": 0, + "fn": 60, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 41, + 59, + 0 + ], + [ + 1, + 59, + 0 + ], + [ + 4, + 56, + 0 + ] + ] + }, + "fra": { + "rows": 15, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 15, + "tp": 15, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 15, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "ja": { + "rows": 220, + "accuracy": 0.44545454545454544, + "macro_f1": 0.3759456749612615, + "macro_f1_present_labels": 0.3759456749612615, + "per_label": { + "AR": { + "support": 100, + "tp": 46, + "fp": 13, + "fn": 54, + "precision": 0.7796610169491526, + "recall": 0.46, + "f1": 0.5786163522012578 + }, + "DIFFUSION": { + "support": 60, + "tp": 49, + "fp": 103, + "fn": 11, + "precision": 0.3223684210526316, + "recall": 0.8166666666666667, + "f1": 0.46226415094339623 + }, + "BOTH": { + "support": 60, + "tp": 3, + "fp": 6, + "fn": 57, + "precision": 0.3333333333333333, + "recall": 0.05, + "f1": 0.08695652173913043 + } + }, + "confusion": [ + [ + 46, + 48, + 6 + ], + [ + 11, + 49, + 0 + ], + [ + 2, + 55, + 3 + ] + ] + }, + "spa": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 256, + "accuracy": 0.41796875, + "macro_f1": 0.3385791459849357, + "macro_f1_present_labels": 0.3385791459849357, + "per_label": { + "AR": { + "support": 112, + "tp": 50, + "fp": 28, + "fn": 62, + "precision": 0.6410256410256411, + "recall": 0.44642857142857145, + "f1": 0.5263157894736842 + }, + "DIFFUSION": { + "support": 72, + "tp": 56, + "fp": 113, + "fn": 16, + "precision": 0.33136094674556216, + "recall": 0.7777777777777778, + "f1": 0.46473029045643155 + }, + "BOTH": { + "support": 72, + "tp": 1, + "fp": 8, + "fn": 71, + "precision": 0.1111111111111111, + "recall": 0.013888888888888888, + "f1": 0.024691358024691357 + } + }, + "confusion": [ + [ + 50, + 55, + 7 + ], + [ + 15, + 56, + 1 + ], + [ + 13, + 58, + 1 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 6, + "fn": 3, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 6, + "tp": 0, + "fp": 0, + "fn": 6, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 3, + 3, + 0 + ] + ] + }, + "short": { + "rows": 1339, + "accuracy": 0.46676624346527257, + "macro_f1": 0.37271747042740394, + "macro_f1_present_labels": 0.37271747042740394, + "per_label": { + "AR": { + "support": 619, + "tp": 282, + "fp": 46, + "fn": 337, + "precision": 0.8597560975609756, + "recall": 0.4555735056542811, + "f1": 0.5955649419218585 + }, + "DIFFUSION": { + "support": 360, + "tp": 339, + "fp": 653, + "fn": 21, + "precision": 0.34173387096774194, + "recall": 0.9416666666666667, + "f1": 0.5014792899408284 + }, + "BOTH": { + "support": 360, + "tp": 4, + "fp": 15, + "fn": 356, + "precision": 0.21052631578947367, + "recall": 0.011111111111111112, + "f1": 0.021108179419525065 + } + }, + "confusion": [ + [ + 282, + 323, + 14 + ], + [ + 20, + 339, + 1 + ], + [ + 26, + 330, + 4 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 8, + "tp": 0, + "fp": 0, + "fn": 8, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 8, + "tp": 0, + "fp": 0, + "fn": 8, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.3333333333333333, + "macro_f1": 0.26666666666666666, + "macro_f1_present_labels": 0.26666666666666666, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 8, + "fn": 4, + "precision": 0.3333333333333333, + "recall": 0.5, + "f1": 0.4 + }, + "BOTH": { + "support": 8, + "tp": 0, + "fp": 0, + "fn": 8, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ], + [ + 4, + 4, + 0 + ] + ] + }, + "unspecified": { + "rows": 1339, + "accuracy": 0.46676624346527257, + "macro_f1": 0.37271747042740394, + "macro_f1_present_labels": 0.37271747042740394, + "per_label": { + "AR": { + "support": 619, + "tp": 282, + "fp": 46, + "fn": 337, + "precision": 0.8597560975609756, + "recall": 0.4555735056542811, + "f1": 0.5955649419218585 + }, + "DIFFUSION": { + "support": 360, + "tp": 339, + "fp": 653, + "fn": 21, + "precision": 0.34173387096774194, + "recall": 0.9416666666666667, + "f1": 0.5014792899408284 + }, + "BOTH": { + "support": 360, + "tp": 4, + "fp": 15, + "fn": 356, + "precision": 0.21052631578947367, + "recall": 0.011111111111111112, + "f1": 0.021108179419525065 + } + }, + "confusion": [ + [ + 282, + 323, + 14 + ], + [ + 20, + 339, + 1 + ], + [ + 26, + 330, + 4 + ] + ] + } + } + }, + "actual_tokens": { + "min": 10, + "max": 32768 + }, + "forward_batches_ms": [ + 6446.7283230042085, + 10.122793028131127, + 9.915609029121697, + 10.60864597093314, + 9.46409406606108, + 9.408487007021904, + 9.18875599745661, + 9.207460912875831, + 9.627168998122215, + 9.453007951378822, + 9.44347190670669, + 9.709913050755858, + 9.246921981684864, + 9.530481998808682, + 9.563550935126841, + 9.204318979755044, + 12.11533194873482, + 10.337620973587036, + 10.059229098260403, + 9.48656804393977, + 9.727524011395872, + 9.305126033723354, + 9.306956082582474, + 9.370358078740537, + 9.624226950109005, + 9.914231020957232, + 9.723542025312781, + 9.962645010091364, + 8.913283003494143, + 8.884695009328425, + 9.641692042350769, + 9.30216396227479, + 9.33746702503413, + 9.300264995545149, + 9.47726599406451, + 9.318343014456332, + 9.37076995614916, + 9.122074930928648, + 9.680890012532473, + 9.092447930015624, + 9.602330974303186, + 9.138197056017816, + 9.636719012632966, + 9.29853692650795, + 9.532617987133563, + 9.35582397505641, + 9.25972405821085, + 9.057466988451779, + 9.71064600162208, + 9.222558932378888, + 9.376573958434165, + 9.44302196148783, + 9.570201043970883, + 9.22214100137353, + 9.490184020251036, + 9.412553044967353, + 9.052847046405077, + 8.90732102561742, + 9.635501075536013, + 9.422325994819403, + 9.483458008617163, + 9.357245056889951, + 9.323348058387637, + 9.352453052997589, + 9.479686967097223, + 9.523665066808462, + 10.155343101359904, + 11.546792928129435, + 10.459608980454504, + 10.044816066510975, + 9.434658102691174, + 9.269515983760357, + 9.521526051685214, + 9.515123907476664, + 9.610714972950518, + 11.145189986564219, + 10.389451985247433, + 9.879508055746555, + 12.027611024677753, + 14.165389002300799, + 18.497899058274925, + 13.323834049515426, + 25.66497807856649, + 10.81712101586163, + 9.532027994282544, + 9.62286302819848, + 9.576256037689745, + 9.407049044966698, + 9.413274936378002, + 9.253080934286118, + 9.664084063842893, + 9.84158890787512, + 9.16440796572715, + 9.031309047713876, + 92.17983298003674, + 94.1256070509553, + 88.64477602764964, + 86.3022570265457, + 89.82986700721085, + 92.94207894708961, + 93.198585906066, + 85.87321802042425, + 88.31606793683022, + 85.75395005755126, + 83.37531797587872, + 85.72892798110843, + 88.96502095740288, + 86.0117319971323, + 86.33086597546935, + 86.55514393467456, + 85.48676606733352, + 85.1146939676255, + 277.01961307320744, + 284.71940592862666, + 283.83382700849324, + 291.9321529334411, + 340.95058403909206, + 273.0076120933518, + 272.6765110855922, + 283.3383050747216, + 269.2251419648528, + 283.6749879643321, + 283.6119490675628, + 278.4766390686855, + 277.30921108741313, + 290.81435909029096, + 280.8271209942177, + 273.62499199807644, + 273.96527596283704, + 268.319983035326, + 1035.654591047205, + 1062.3674780363217, + 1062.9829469835386, + 1018.123218906112, + 1085.1643879432231, + 1091.3887589704245, + 1005.2773610223085, + 1050.7057759677991, + 1105.7434329995885, + 1062.8094230778515, + 1055.2198939258233, + 1032.5770160416141, + 1053.3472939860076, + 1077.022343990393, + 1085.838780971244, + 1035.0606159772724, + 1080.3039139136672, + 1039.5317010115832, + 3934.3111380003393, + 3868.6304059810936, + 4303.185440949164, + 4105.504778912291, + 4010.3370539145544, + 3995.786834973842, + 4333.715492975898, + 4184.374890988693, + 4037.6546720508486, + 4065.9116219030693, + 4104.565493064001, + 4185.04842906259, + 4191.750784986652, + 4087.02986210119, + 4069.8690769495443, + 4116.871495964006, + 4112.496968009509, + 4220.928212045692 + ], + "precision": { + "parameters": "float32", + "forward_autocast": null, + "device_type": "cuda" + }, + "data_files": [ + { + "file": "test.jsonl", + "sha256": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53" + }, + { + "file": "test.jsonl", + "sha256": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c" + } + ] +} diff --git a/evaluation/final-baseline-fp32.predictions.jsonl b/evaluation/final-baseline-fp32.predictions.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..da43f6845594b0ca09e42e5183fb50909c2ca5dd --- /dev/null +++ b/evaluation/final-baseline-fp32.predictions.jsonl @@ -0,0 +1,1411 @@ +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-AR", "prediction": 1, "probabilities": [0.12958167493343353, 0.8528487086296082, 0.017569635063409805], "tokens": 39} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-DIFFUSION", "prediction": 1, "probabilities": [0.09633228182792664, 0.889589250087738, 0.014078467153012753], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-BOTH", "prediction": 1, "probabilities": [0.09578964114189148, 0.8884662985801697, 0.015744047239422798], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-AR", "prediction": 1, "probabilities": [0.15650127828121185, 0.7324252128601074, 0.11107347160577774], "tokens": 38} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-DIFFUSION", "prediction": 1, "probabilities": [0.22842095792293549, 0.6788163781166077, 0.09276262670755386], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-BOTH", "prediction": 1, "probabilities": [0.092164546251297, 0.8662823438644409, 0.04155304655432701], "tokens": 33} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-AR", "prediction": 1, "probabilities": [0.19788792729377747, 0.7861651182174683, 0.015947027131915092], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-DIFFUSION", "prediction": 1, "probabilities": [0.17463767528533936, 0.8114065527915955, 0.013955761678516865], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-BOTH", "prediction": 1, "probabilities": [0.12451109290122986, 0.8667944669723511, 0.008694469928741455], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-AR", "prediction": 1, "probabilities": [0.07563791424036026, 0.9200043678283691, 0.004357724450528622], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1416817605495453, 0.849981963634491, 0.008336273021996021], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-BOTH", "prediction": 1, "probabilities": [0.10618505626916885, 0.887681245803833, 0.006133695133030415], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-AR", "prediction": 1, "probabilities": [0.28560975193977356, 0.6945858597755432, 0.019804347306489944], "tokens": 38} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-DIFFUSION", "prediction": 1, "probabilities": [0.2828160524368286, 0.7076544761657715, 0.009529485367238522], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-BOTH", "prediction": 1, "probabilities": [0.3921014070510864, 0.5892018675804138, 0.018696710467338562], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-AR", "prediction": 1, "probabilities": [0.2888640761375427, 0.6244733333587646, 0.08666260540485382], "tokens": 39} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-DIFFUSION", "prediction": 1, "probabilities": [0.31397297978401184, 0.64869624376297, 0.03733077272772789], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-BOTH", "prediction": 1, "probabilities": [0.09070304036140442, 0.8525505661964417, 0.056746359914541245], "tokens": 35} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-AR", "prediction": 1, "probabilities": [0.042938634753227234, 0.9334527254104614, 0.023608660325407982], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-DIFFUSION", "prediction": 1, "probabilities": [0.02374276891350746, 0.9596847295761108, 0.016572481021285057], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-BOTH", "prediction": 1, "probabilities": [0.029228420928120613, 0.9438303112983704, 0.02694130316376686], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-AR", "prediction": 1, "probabilities": [0.07651384174823761, 0.8751413822174072, 0.04834483191370964], "tokens": 41} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07965601980686188, 0.8716477155685425, 0.04869619756937027], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-BOTH", "prediction": 1, "probabilities": [0.07095590978860855, 0.8650882840156555, 0.06395582109689713], "tokens": 40} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-AR", "prediction": 1, "probabilities": [0.09546825289726257, 0.887227475643158, 0.017304271459579468], "tokens": 46} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-DIFFUSION", "prediction": 1, "probabilities": [0.05889270827174187, 0.9217090606689453, 0.019398167729377747], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-BOTH", "prediction": 1, "probabilities": [0.06402722746133804, 0.8941044807434082, 0.04186832159757614], "tokens": 43} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-AR", "prediction": 1, "probabilities": [0.07877219468355179, 0.907717227935791, 0.013510552234947681], "tokens": 47} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-DIFFUSION", "prediction": 1, "probabilities": [0.055849917232990265, 0.9204328656196594, 0.023717252537608147], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-BOTH", "prediction": 1, "probabilities": [0.06622689962387085, 0.909163773059845, 0.024609336629509926], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-AR", "prediction": 1, "probabilities": [0.30986806750297546, 0.6541123986244202, 0.036019522696733475], "tokens": 46} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-DIFFUSION", "prediction": 1, "probabilities": [0.06229597330093384, 0.9104943871498108, 0.027209650725126266], "tokens": 40} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-BOTH", "prediction": 1, "probabilities": [0.26430195569992065, 0.6893665790557861, 0.046331461519002914], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-AR", "prediction": 1, "probabilities": [0.1628509759902954, 0.7418752908706665, 0.09527377784252167], "tokens": 47} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-DIFFUSION", "prediction": 1, "probabilities": [0.12655703723430634, 0.7725074887275696, 0.1009354293346405], "tokens": 43} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-BOTH", "prediction": 1, "probabilities": [0.08625024557113647, 0.824271023273468, 0.0894787386059761], "tokens": 41} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-AR", "prediction": 1, "probabilities": [0.23093873262405396, 0.7407177686691284, 0.02834351360797882], "tokens": 76} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-DIFFUSION", "prediction": 1, "probabilities": [0.24385350942611694, 0.7298979759216309, 0.026248561218380928], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-BOTH", "prediction": 1, "probabilities": [0.24093957245349884, 0.725807785987854, 0.03325260803103447], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-AR", "prediction": 1, "probabilities": [0.25090649724006653, 0.6791988611221313, 0.06989465653896332], "tokens": 75} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-DIFFUSION", "prediction": 1, "probabilities": [0.34140077233314514, 0.6207175254821777, 0.03788170963525772], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH", "prediction": 1, "probabilities": [0.21519897878170013, 0.7515274882316589, 0.03327351063489914], "tokens": 70} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-AR", "prediction": 1, "probabilities": [0.22222870588302612, 0.753422737121582, 0.0243485439568758], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-DIFFUSION", "prediction": 1, "probabilities": [0.25410425662994385, 0.7242043614387512, 0.021691445261240005], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-BOTH", "prediction": 1, "probabilities": [0.18428190052509308, 0.7973511815071106, 0.018366973847150803], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-AR", "prediction": 1, "probabilities": [0.19377945363521576, 0.7958406209945679, 0.010379915125668049], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-DIFFUSION", "prediction": 1, "probabilities": [0.31075817346572876, 0.6747374534606934, 0.01450437493622303], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-BOTH", "prediction": 1, "probabilities": [0.20445430278778076, 0.7810729742050171, 0.014472730457782745], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-AR", "prediction": 1, "probabilities": [0.3378753662109375, 0.6417402625083923, 0.020384375005960464], "tokens": 75} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-DIFFUSION", "prediction": 1, "probabilities": [0.4334220588207245, 0.5521525740623474, 0.014425359666347504], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-BOTH", "prediction": 1, "probabilities": [0.3629038333892822, 0.6113210916519165, 0.025775110349059105], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-AR", "prediction": 1, "probabilities": [0.29965949058532715, 0.6521914601325989, 0.048149023205041885], "tokens": 76} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-DIFFUSION", "prediction": 1, "probabilities": [0.34260740876197815, 0.6278369426727295, 0.029555637389421463], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-BOTH", "prediction": 1, "probabilities": [0.16895678639411926, 0.7914445400238037, 0.039598651230335236], "tokens": 72} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-AR", "prediction": 1, "probabilities": [0.026704514399170876, 0.959112286567688, 0.014183170162141323], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-DIFFUSION", "prediction": 1, "probabilities": [0.012113431468605995, 0.9815071821212769, 0.006379446946084499], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-BOTH", "prediction": 1, "probabilities": [0.022881140932440758, 0.96341472864151, 0.013704170472919941], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-AR", "prediction": 1, "probabilities": [0.10250361263751984, 0.8351184725761414, 0.0623779334127903], "tokens": 30} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-DIFFUSION", "prediction": 1, "probabilities": [0.06927783042192459, 0.9054103493690491, 0.02531183883547783], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-BOTH", "prediction": 1, "probabilities": [0.07662032544612885, 0.8767461180686951, 0.046633537858724594], "tokens": 29} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-AR", "prediction": 1, "probabilities": [0.1727643758058548, 0.8187238574028015, 0.00851172860711813], "tokens": 35} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-DIFFUSION", "prediction": 1, "probabilities": [0.0658629983663559, 0.9279698729515076, 0.0061671850271523], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-BOTH", "prediction": 1, "probabilities": [0.10771384835243225, 0.8738940954208374, 0.018392033874988556], "tokens": 32} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-AR", "prediction": 1, "probabilities": [0.09541347622871399, 0.8989087343215942, 0.005677764769643545], "tokens": 36} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-DIFFUSION", "prediction": 1, "probabilities": [0.0745665580034256, 0.9183447360992432, 0.0070886798202991486], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-BOTH", "prediction": 1, "probabilities": [0.134994775056839, 0.8555949330329895, 0.009410310536623001], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-AR", "prediction": 0, "probabilities": [0.5269973874092102, 0.4520761966705322, 0.02092638798058033], "tokens": 35} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-DIFFUSION", "prediction": 1, "probabilities": [0.07129212468862534, 0.9199647307395935, 0.008743202313780785], "tokens": 29} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-BOTH", "prediction": 0, "probabilities": [0.5806640386581421, 0.4026433527469635, 0.01669257879257202], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-AR", "prediction": 1, "probabilities": [0.30383116006851196, 0.6211891174316406, 0.07497971504926682], "tokens": 36} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-DIFFUSION", "prediction": 1, "probabilities": [0.20231367647647858, 0.7413339018821716, 0.0563524104654789], "tokens": 32} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-BOTH", "prediction": 1, "probabilities": [0.1534455418586731, 0.7778611183166504, 0.0686933621764183], "tokens": 30} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-AR", "prediction": 1, "probabilities": [0.044684309512376785, 0.9405511021614075, 0.01476452499628067], "tokens": 63} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-DIFFUSION", "prediction": 1, "probabilities": [0.026418261229991913, 0.9601250886917114, 0.01345669012516737], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-BOTH", "prediction": 1, "probabilities": [0.030842259526252747, 0.9532838463783264, 0.01587386429309845], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-AR", "prediction": 1, "probabilities": [0.047017402946949005, 0.9103559851646423, 0.04262659698724747], "tokens": 62} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-DIFFUSION", "prediction": 1, "probabilities": [0.02886282093822956, 0.9465508460998535, 0.024586383253335953], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-BOTH", "prediction": 1, "probabilities": [0.030623596161603928, 0.949458122253418, 0.019918221980333328], "tokens": 57} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-AR", "prediction": 1, "probabilities": [0.040381841361522675, 0.9479458332061768, 0.011672348715364933], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-DIFFUSION", "prediction": 1, "probabilities": [0.028500529006123543, 0.9577072858810425, 0.013792233541607857], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-BOTH", "prediction": 1, "probabilities": [0.026474473997950554, 0.9659964442253113, 0.007529108785092831], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-AR", "prediction": 1, "probabilities": [0.02116158790886402, 0.9731072783470154, 0.005731193348765373], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-DIFFUSION", "prediction": 1, "probabilities": [0.02984045073390007, 0.9603596329689026, 0.009799858555197716], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-BOTH", "prediction": 1, "probabilities": [0.02678731456398964, 0.964287281036377, 0.008925438858568668], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-AR", "prediction": 1, "probabilities": [0.061613570898771286, 0.9248413443565369, 0.013545078225433826], "tokens": 62} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-DIFFUSION", "prediction": 1, "probabilities": [0.04665450006723404, 0.9419063925743103, 0.01143905520439148], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-BOTH", "prediction": 1, "probabilities": [0.060170214623212814, 0.9230188131332397, 0.016810975968837738], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-AR", "prediction": 1, "probabilities": [0.04883717745542526, 0.9104605317115784, 0.04070233181118965], "tokens": 63} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-DIFFUSION", "prediction": 1, "probabilities": [0.05962255969643593, 0.9132973551750183, 0.027080029249191284], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-BOTH", "prediction": 1, "probabilities": [0.02385900914669037, 0.9509842991828918, 0.025156617164611816], "tokens": 59} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-AR", "prediction": 1, "probabilities": [0.11362401396036148, 0.8610513210296631, 0.025324713438749313], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-DIFFUSION", "prediction": 1, "probabilities": [0.05510544031858444, 0.9283276200294495, 0.01656688190996647], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-BOTH", "prediction": 1, "probabilities": [0.11290835589170456, 0.8551754951477051, 0.031916193664073944], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-AR", "prediction": 1, "probabilities": [0.2050783485174179, 0.7446529865264893, 0.05026872828602791], "tokens": 48} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-DIFFUSION", "prediction": 1, "probabilities": [0.1871626079082489, 0.7846968173980713, 0.028140567243099213], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-BOTH", "prediction": 1, "probabilities": [0.15102653205394745, 0.8049753904342651, 0.04399808496236801], "tokens": 47} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-AR", "prediction": 1, "probabilities": [0.1936911940574646, 0.7874675393104553, 0.018841227516531944], "tokens": 53} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-DIFFUSION", "prediction": 1, "probabilities": [0.09783370792865753, 0.8891408443450928, 0.013025468215346336], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-BOTH", "prediction": 1, "probabilities": [0.11373980343341827, 0.8622010946273804, 0.02405904233455658], "tokens": 50} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-AR", "prediction": 1, "probabilities": [0.2076587975025177, 0.7790408134460449, 0.013300403021275997], "tokens": 54} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-DIFFUSION", "prediction": 1, "probabilities": [0.19737720489501953, 0.7841629385948181, 0.018459903076291084], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-BOTH", "prediction": 1, "probabilities": [0.1954311579465866, 0.7845119833946228, 0.0200568325817585], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-AR", "prediction": 1, "probabilities": [0.4599570631980896, 0.5124526023864746, 0.027590380981564522], "tokens": 53} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-DIFFUSION", "prediction": 1, "probabilities": [0.201478511095047, 0.7708261609077454, 0.027695389464497566], "tokens": 47} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-BOTH", "prediction": 1, "probabilities": [0.4557740390300751, 0.5207744836807251, 0.023451512679457664], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-AR", "prediction": 1, "probabilities": [0.36009928584098816, 0.5704972743988037, 0.06940344721078873], "tokens": 54} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-DIFFUSION", "prediction": 1, "probabilities": [0.2403634786605835, 0.7110819220542908, 0.04855461046099663], "tokens": 50} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-BOTH", "prediction": 1, "probabilities": [0.20737288892269135, 0.7371106147766113, 0.05551646277308464], "tokens": 48} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-AR", "prediction": 1, "probabilities": [0.1945399045944214, 0.6912226676940918, 0.11423742771148682], "tokens": 38} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-DIFFUSION", "prediction": 1, "probabilities": [0.15846887230873108, 0.7123583555221558, 0.12917275726795197], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-BOTH", "prediction": 1, "probabilities": [0.1949373334646225, 0.6037163138389587, 0.20134638249874115], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-AR", "prediction": 1, "probabilities": [0.18254335224628448, 0.4992956817150116, 0.3181609809398651], "tokens": 37} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-DIFFUSION", "prediction": 1, "probabilities": [0.25179558992385864, 0.4438076317310333, 0.30439671874046326], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-BOTH", "prediction": 1, "probabilities": [0.13548296689987183, 0.7131534814834595, 0.1513635665178299], "tokens": 32} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-AR", "prediction": 1, "probabilities": [0.24840474128723145, 0.6796287894248962, 0.07196643203496933], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-DIFFUSION", "prediction": 1, "probabilities": [0.22072435915470123, 0.6437357664108276, 0.13553984463214874], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-BOTH", "prediction": 1, "probabilities": [0.17739860713481903, 0.7715349197387695, 0.05106651782989502], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-AR", "prediction": 1, "probabilities": [0.1479281187057495, 0.8270103931427002, 0.025061529129743576], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-DIFFUSION", "prediction": 1, "probabilities": [0.19702613353729248, 0.7163659930229187, 0.08660788834095001], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-BOTH", "prediction": 1, "probabilities": [0.13368791341781616, 0.7956284880638123, 0.07068359106779099], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-AR", "prediction": 1, "probabilities": [0.40566977858543396, 0.5100123882293701, 0.08431781828403473], "tokens": 37} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-DIFFUSION", "prediction": 1, "probabilities": [0.4079194962978363, 0.5125123262405396, 0.07956816256046295], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-BOTH", "prediction": 1, "probabilities": [0.4115535318851471, 0.4340522885322571, 0.15439417958259583], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-AR", "prediction": 2, "probabilities": [0.31321194767951965, 0.32582178711891174, 0.36096620559692383], "tokens": 38} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-DIFFUSION", "prediction": 0, "probabilities": [0.4578494131565094, 0.36526429653167725, 0.17688633501529694], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-BOTH", "prediction": 1, "probabilities": [0.11180663853883743, 0.592430055141449, 0.2957633435726166], "tokens": 34} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-AR", "prediction": 1, "probabilities": [0.14476217329502106, 0.8290627002716064, 0.02617514878511429], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-DIFFUSION", "prediction": 1, "probabilities": [0.07189334183931351, 0.9104292988777161, 0.01767732948064804], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-BOTH", "prediction": 1, "probabilities": [0.16599641740322113, 0.7914891242980957, 0.04251445457339287], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-AR", "prediction": 0, "probabilities": [0.46661877632141113, 0.44681382179260254, 0.08656742423772812], "tokens": 26} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-DIFFUSION", "prediction": 1, "probabilities": [0.3265221416950226, 0.620927631855011, 0.05255019664764404], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-BOTH", "prediction": 0, "probabilities": [0.45149120688438416, 0.4349002242088318, 0.11360851675271988], "tokens": 25} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-AR", "prediction": 0, "probabilities": [0.6397732496261597, 0.34985673427581787, 0.010369976051151752], "tokens": 31} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-DIFFUSION", "prediction": 1, "probabilities": [0.374033659696579, 0.6132009029388428, 0.012765452265739441], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-BOTH", "prediction": 1, "probabilities": [0.380939781665802, 0.5511927604675293, 0.06786748766899109], "tokens": 28} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-AR", "prediction": 1, "probabilities": [0.47749748826026917, 0.5145114660263062, 0.007991032674908638], "tokens": 32} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-DIFFUSION", "prediction": 0, "probabilities": [0.5186256170272827, 0.46661415696144104, 0.014760171994566917], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-BOTH", "prediction": 0, "probabilities": [0.6767075061798096, 0.30581995844841003, 0.017472608014941216], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-AR", "prediction": 0, "probabilities": [0.9303337335586548, 0.059070661664009094, 0.010595632717013359], "tokens": 31} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-DIFFUSION", "prediction": 1, "probabilities": [0.3987230062484741, 0.5854468941688538, 0.015830116346478462], "tokens": 25} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-BOTH", "prediction": 0, "probabilities": [0.9719512462615967, 0.02185453660786152, 0.006194224115461111], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-AR", "prediction": 0, "probabilities": [0.7680595517158508, 0.1649668663740158, 0.06697352975606918], "tokens": 32} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-DIFFUSION", "prediction": 0, "probabilities": [0.7881633639335632, 0.15972256660461426, 0.052114006131887436], "tokens": 28} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-BOTH", "prediction": 0, "probabilities": [0.5679279565811157, 0.28766775131225586, 0.14440429210662842], "tokens": 26} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-AR", "prediction": 1, "probabilities": [0.050485167652368546, 0.9374814033508301, 0.012033458799123764], "tokens": 44} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-DIFFUSION", "prediction": 1, "probabilities": [0.05540524795651436, 0.9357365965843201, 0.00885818898677826], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-BOTH", "prediction": 1, "probabilities": [0.03895110264420509, 0.9483917951583862, 0.012657089158892632], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-AR", "prediction": 1, "probabilities": [0.06881489604711533, 0.8580036163330078, 0.07318145036697388], "tokens": 43} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION", "prediction": 1, "probabilities": [0.1163879930973053, 0.8326637148857117, 0.050948336720466614], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-BOTH", "prediction": 1, "probabilities": [0.04938982427120209, 0.9227674007415771, 0.027842741459608078], "tokens": 38} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-AR", "prediction": 1, "probabilities": [0.05711088702082634, 0.9333008527755737, 0.009588271379470825], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-DIFFUSION", "prediction": 1, "probabilities": [0.06948041915893555, 0.9179106950759888, 0.01260883267968893], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-BOTH", "prediction": 1, "probabilities": [0.05747705698013306, 0.9336138367652893, 0.008909063413739204], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-AR", "prediction": 1, "probabilities": [0.023890601471066475, 0.9733027219772339, 0.002806631615385413], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-DIFFUSION", "prediction": 1, "probabilities": [0.05070502310991287, 0.9412022233009338, 0.008092754520475864], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-BOTH", "prediction": 1, "probabilities": [0.043503955006599426, 0.949344277381897, 0.007151791360229254], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-AR", "prediction": 1, "probabilities": [0.0933249443769455, 0.893581211566925, 0.013093795627355576], "tokens": 43} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-DIFFUSION", "prediction": 1, "probabilities": [0.09655685722827911, 0.8958885669708252, 0.0075545161962509155], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-BOTH", "prediction": 1, "probabilities": [0.13138200342655182, 0.8533668518066406, 0.015251155942678452], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-AR", "prediction": 1, "probabilities": [0.09465335309505463, 0.8468688130378723, 0.05847791209816933], "tokens": 44} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-DIFFUSION", "prediction": 1, "probabilities": [0.1294696033000946, 0.8429557681083679, 0.027574589475989342], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-BOTH", "prediction": 1, "probabilities": [0.035417456179857254, 0.9176353216171265, 0.046947259455919266], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-AR", "prediction": 1, "probabilities": [0.09132761508226395, 0.8917251229286194, 0.016947263851761818], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-DIFFUSION", "prediction": 1, "probabilities": [0.08030220121145248, 0.9080915451049805, 0.011606295593082905], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-BOTH", "prediction": 1, "probabilities": [0.13324512541294098, 0.848133385181427, 0.018621468916535378], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-AR", "prediction": 1, "probabilities": [0.2243415266275406, 0.7265411615371704, 0.04911725968122482], "tokens": 38} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2509298324584961, 0.7160834670066833, 0.032986678183078766], "tokens": 36} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-BOTH", "prediction": 1, "probabilities": [0.22721520066261292, 0.719081461429596, 0.053703345358371735], "tokens": 37} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-AR", "prediction": 1, "probabilities": [0.34777188301086426, 0.6425108909606934, 0.009717287495732307], "tokens": 42} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-DIFFUSION", "prediction": 1, "probabilities": [0.19396886229515076, 0.7955671548843384, 0.010463962331414223], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-BOTH", "prediction": 1, "probabilities": [0.3382524847984314, 0.6332166790962219, 0.028530795127153397], "tokens": 39} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-AR", "prediction": 1, "probabilities": [0.18276171386241913, 0.8087491393089294, 0.00848917942494154], "tokens": 43} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1635616272687912, 0.8230164051055908, 0.01342197135090828], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-BOTH", "prediction": 1, "probabilities": [0.32684987783432007, 0.659248948097229, 0.013901199214160442], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-AR", "prediction": 0, "probabilities": [0.6429681181907654, 0.34100398421287537, 0.016027912497520447], "tokens": 42} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-DIFFUSION", "prediction": 1, "probabilities": [0.28044477105140686, 0.7032222747802734, 0.016332928091287613], "tokens": 36} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-BOTH", "prediction": 0, "probabilities": [0.7842684984207153, 0.2033877819776535, 0.01234367024153471], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-AR", "prediction": 1, "probabilities": [0.4375612139701843, 0.501211404800415, 0.06122738495469093], "tokens": 44} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-DIFFUSION", "prediction": 1, "probabilities": [0.39948147535324097, 0.5403581857681274, 0.06016034632921219], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-BOTH", "prediction": 1, "probabilities": [0.2990701496601105, 0.6366137266159058, 0.06431618332862854], "tokens": 38} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-AR", "prediction": 1, "probabilities": [0.05587827041745186, 0.9375885128974915, 0.006533192470669746], "tokens": 71} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-DIFFUSION", "prediction": 1, "probabilities": [0.044486504048109055, 0.9507380723953247, 0.004775484558194876], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-BOTH", "prediction": 1, "probabilities": [0.045354731380939484, 0.9488977193832397, 0.005747511982917786], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-AR", "prediction": 1, "probabilities": [0.06588021665811539, 0.915987491607666, 0.018132315948605537], "tokens": 70} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-DIFFUSION", "prediction": 1, "probabilities": [0.05331292003393173, 0.9387755393981934, 0.007911511696875095], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-BOTH", "prediction": 1, "probabilities": [0.05041651800274849, 0.9423062205314636, 0.007277211174368858], "tokens": 65} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-AR", "prediction": 1, "probabilities": [0.05494684725999832, 0.9396557807922363, 0.005397322122007608], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-DIFFUSION", "prediction": 1, "probabilities": [0.04574966803193092, 0.9497203230857849, 0.0045300195924937725], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-BOTH", "prediction": 1, "probabilities": [0.047775428742170334, 0.9484813213348389, 0.0037432305980473757], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-AR", "prediction": 1, "probabilities": [0.0378696471452713, 0.9594204425811768, 0.0027099011931568384], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-DIFFUSION", "prediction": 1, "probabilities": [0.04842664301395416, 0.9481779336929321, 0.0033954179380089045], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-BOTH", "prediction": 1, "probabilities": [0.03889194503426552, 0.9575448632240295, 0.0035632196813821793], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-AR", "prediction": 1, "probabilities": [0.07554520666599274, 0.9192458987236023, 0.005208937916904688], "tokens": 70} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-DIFFUSION", "prediction": 1, "probabilities": [0.07502246648073196, 0.9206973314285278, 0.004280195105820894], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-BOTH", "prediction": 1, "probabilities": [0.07413571327924728, 0.9194403886795044, 0.0064238994382321835], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-AR", "prediction": 1, "probabilities": [0.0653238296508789, 0.9220759272575378, 0.012600217945873737], "tokens": 71} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-DIFFUSION", "prediction": 1, "probabilities": [0.0817038044333458, 0.9107791185379028, 0.007517105899751186], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-BOTH", "prediction": 1, "probabilities": [0.037637338042259216, 0.9542985558509827, 0.008064120076596737], "tokens": 67} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-AR", "prediction": 1, "probabilities": [0.015154466032981873, 0.9783312082290649, 0.006514288019388914], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-DIFFUSION", "prediction": 1, "probabilities": [0.013102009892463684, 0.9825673699378967, 0.004330577328801155], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-BOTH", "prediction": 1, "probabilities": [0.028825122863054276, 0.9636317491531372, 0.007543101906776428], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-AR", "prediction": 1, "probabilities": [0.014411461539566517, 0.9729486107826233, 0.01263998169451952], "tokens": 77} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-DIFFUSION", "prediction": 1, "probabilities": [0.018212608993053436, 0.9726357460021973, 0.009151604026556015], "tokens": 75} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-BOTH", "prediction": 1, "probabilities": [0.01936681754887104, 0.9685738682746887, 0.012059344910085201], "tokens": 76} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-AR", "prediction": 1, "probabilities": [0.026155034080147743, 0.9662739634513855, 0.007570994086563587], "tokens": 81} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-DIFFUSION", "prediction": 1, "probabilities": [0.02342422492802143, 0.9714555740356445, 0.005120211746543646], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-BOTH", "prediction": 1, "probabilities": [0.032095443457365036, 0.9589316248893738, 0.00897290837019682], "tokens": 78} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-AR", "prediction": 1, "probabilities": [0.02471771463751793, 0.9692423343658447, 0.006039976142346859], "tokens": 82} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-DIFFUSION", "prediction": 1, "probabilities": [0.017575152218341827, 0.9777960181236267, 0.004628795199096203], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-BOTH", "prediction": 1, "probabilities": [0.040700122714042664, 0.9543951153755188, 0.004904804285615683], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-AR", "prediction": 1, "probabilities": [0.06569956243038177, 0.9215298891067505, 0.012770604342222214], "tokens": 81} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-DIFFUSION", "prediction": 1, "probabilities": [0.02827112376689911, 0.964823842048645, 0.006904997397214174], "tokens": 75} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-BOTH", "prediction": 1, "probabilities": [0.06843669712543488, 0.9226399064064026, 0.008923406712710857], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-AR", "prediction": 1, "probabilities": [0.03077968955039978, 0.9552862048149109, 0.013934117741882801], "tokens": 83} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-DIFFUSION", "prediction": 1, "probabilities": [0.02886880561709404, 0.9573301076889038, 0.013801028952002525], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-BOTH", "prediction": 1, "probabilities": [0.022369252517819405, 0.9651991128921509, 0.012431617826223373], "tokens": 77} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-AR", "prediction": 1, "probabilities": [0.04013883322477341, 0.9460926651954651, 0.013768467120826244], "tokens": 55} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-DIFFUSION", "prediction": 1, "probabilities": [0.02565382793545723, 0.9629902243614197, 0.011355890892446041], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-BOTH", "prediction": 1, "probabilities": [0.029467537999153137, 0.9585636258125305, 0.011968802660703659], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-AR", "prediction": 1, "probabilities": [0.036771032959222794, 0.9300064444541931, 0.033222496509552], "tokens": 54} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.03596087172627449, 0.9373319149017334, 0.026707231998443604], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-BOTH", "prediction": 1, "probabilities": [0.031194018200039864, 0.9504311680793762, 0.018374867737293243], "tokens": 49} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-AR", "prediction": 1, "probabilities": [0.03254767507314682, 0.9550514817237854, 0.01240085531026125], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-DIFFUSION", "prediction": 1, "probabilities": [0.03289090842008591, 0.9535889029502869, 0.01352024544030428], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-BOTH", "prediction": 1, "probabilities": [0.027931343764066696, 0.9628131985664368, 0.009255509823560715], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-AR", "prediction": 1, "probabilities": [0.01939770020544529, 0.9764665961265564, 0.0041356682777404785], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-DIFFUSION", "prediction": 1, "probabilities": [0.026761198416352272, 0.9638637900352478, 0.009374977089464664], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-BOTH", "prediction": 1, "probabilities": [0.01956341229379177, 0.9729318618774414, 0.007504737935960293], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-AR", "prediction": 1, "probabilities": [0.054121650755405426, 0.9317250847816467, 0.01415320672094822], "tokens": 54} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-DIFFUSION", "prediction": 1, "probabilities": [0.04619155079126358, 0.9435796737670898, 0.010228783823549747], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-BOTH", "prediction": 1, "probabilities": [0.05113936588168144, 0.9320387840270996, 0.016821861267089844], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-AR", "prediction": 1, "probabilities": [0.04400433972477913, 0.9156258702278137, 0.04036983475089073], "tokens": 55} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-DIFFUSION", "prediction": 1, "probabilities": [0.04745978116989136, 0.9240535497665405, 0.028486639261245728], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-BOTH", "prediction": 1, "probabilities": [0.023145051673054695, 0.9445908069610596, 0.03226413577795029], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-AR", "prediction": 1, "probabilities": [0.02148349955677986, 0.9701827168464661, 0.008333834819495678], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-DIFFUSION", "prediction": 1, "probabilities": [0.012899789959192276, 0.98395174741745, 0.0031484125647693872], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-BOTH", "prediction": 1, "probabilities": [0.024674661457538605, 0.969331681728363, 0.005993660073727369], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-AR", "prediction": 1, "probabilities": [0.022860730066895485, 0.9657652378082275, 0.01137405913323164], "tokens": 54} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-DIFFUSION", "prediction": 1, "probabilities": [0.01762847788631916, 0.9766942262649536, 0.005677256267517805], "tokens": 52} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-BOTH", "prediction": 1, "probabilities": [0.029820119962096214, 0.9558853507041931, 0.014294487424194813], "tokens": 53} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-AR", "prediction": 1, "probabilities": [0.04152638837695122, 0.9525658488273621, 0.005907754879444838], "tokens": 58} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-DIFFUSION", "prediction": 1, "probabilities": [0.024465912953019142, 0.97287917137146, 0.0026549454778432846], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-BOTH", "prediction": 1, "probabilities": [0.03487701714038849, 0.9589188694953918, 0.006204136181622744], "tokens": 55} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-AR", "prediction": 1, "probabilities": [0.04001818224787712, 0.9558137655258179, 0.004168021958321333], "tokens": 59} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-DIFFUSION", "prediction": 1, "probabilities": [0.022971592843532562, 0.9733867645263672, 0.003641608403995633], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-BOTH", "prediction": 1, "probabilities": [0.06310809403657913, 0.9318228960037231, 0.005069045815616846], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-AR", "prediction": 1, "probabilities": [0.11784624308347702, 0.8731310367584229, 0.009022744372487068], "tokens": 58} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-DIFFUSION", "prediction": 1, "probabilities": [0.034517496824264526, 0.9610311388969421, 0.004451339598745108], "tokens": 52} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-BOTH", "prediction": 1, "probabilities": [0.11644871532917023, 0.8763439655303955, 0.007207377813756466], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-AR", "prediction": 1, "probabilities": [0.06564319133758545, 0.9185624718666077, 0.01579434424638748], "tokens": 60} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-DIFFUSION", "prediction": 1, "probabilities": [0.04221871867775917, 0.9442406296730042, 0.013540721498429775], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-BOTH", "prediction": 1, "probabilities": [0.03629123792052269, 0.9480561017990112, 0.015652624890208244], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-AR", "prediction": 1, "probabilities": [0.018248163163661957, 0.9745495915412903, 0.0072022611275315285], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-DIFFUSION", "prediction": 1, "probabilities": [0.014893818646669388, 0.9801622629165649, 0.004943909589201212], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-BOTH", "prediction": 1, "probabilities": [0.010528531856834888, 0.9846220016479492, 0.004849523305892944], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-AR", "prediction": 1, "probabilities": [0.020501255989074707, 0.9591787457466125, 0.020319974049925804], "tokens": 53} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-DIFFUSION", "prediction": 1, "probabilities": [0.017568588256835938, 0.9704632759094238, 0.011968100443482399], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-BOTH", "prediction": 1, "probabilities": [0.013404777273535728, 0.9782335758209229, 0.008361642248928547], "tokens": 48} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-AR", "prediction": 1, "probabilities": [0.019339988008141518, 0.9758713841438293, 0.004788626451045275], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-DIFFUSION", "prediction": 1, "probabilities": [0.01919196918606758, 0.9762482047080994, 0.004559841938316822], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-BOTH", "prediction": 1, "probabilities": [0.014906339347362518, 0.9823933839797974, 0.0027002242859452963], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-AR", "prediction": 1, "probabilities": [0.00869203731417656, 0.9897546172142029, 0.001553372130729258], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-DIFFUSION", "prediction": 1, "probabilities": [0.01306823268532753, 0.9841229319572449, 0.00280885212123394], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-BOTH", "prediction": 1, "probabilities": [0.010285048745572567, 0.9872258305549622, 0.0024891230277717113], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-AR", "prediction": 1, "probabilities": [0.03142944350838661, 0.9638491868972778, 0.00472137751057744], "tokens": 53} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-DIFFUSION", "prediction": 1, "probabilities": [0.023825425654649734, 0.9734371304512024, 0.0027373989578336477], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-BOTH", "prediction": 1, "probabilities": [0.02989235334098339, 0.9642956256866455, 0.0058120558969676495], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-AR", "prediction": 1, "probabilities": [0.019013188779354095, 0.9659885764122009, 0.014998222701251507], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-DIFFUSION", "prediction": 1, "probabilities": [0.026054395362734795, 0.9645811319351196, 0.009364486671984196], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-BOTH", "prediction": 1, "probabilities": [0.010083973407745361, 0.977469265460968, 0.012446749955415726], "tokens": 50} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-AR", "prediction": 1, "probabilities": [0.02921418286859989, 0.9655101299285889, 0.005275629926472902], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-DIFFUSION", "prediction": 1, "probabilities": [0.024306850507855415, 0.9722107648849487, 0.003482334315776825], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-BOTH", "prediction": 1, "probabilities": [0.030944472178816795, 0.9637967944145203, 0.005258677527308464], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-AR", "prediction": 1, "probabilities": [0.06895128637552261, 0.9204940795898438, 0.010554653592407703], "tokens": 53} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-DIFFUSION", "prediction": 1, "probabilities": [0.049577340483665466, 0.9435446262359619, 0.006878003478050232], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-BOTH", "prediction": 1, "probabilities": [0.07284076511859894, 0.9133896231651306, 0.013769573532044888], "tokens": 52} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-AR", "prediction": 1, "probabilities": [0.07024337351322174, 0.9247878193855286, 0.004968788009136915], "tokens": 58} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-DIFFUSION", "prediction": 1, "probabilities": [0.03756420314311981, 0.9588143229484558, 0.0036215255968272686], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-BOTH", "prediction": 1, "probabilities": [0.06456363946199417, 0.926981508731842, 0.0084548844024539], "tokens": 55} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-AR", "prediction": 1, "probabilities": [0.06299073249101639, 0.9336211085319519, 0.0033881429117172956], "tokens": 59} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-DIFFUSION", "prediction": 1, "probabilities": [0.052674662321805954, 0.9430506229400635, 0.004274725914001465], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-BOTH", "prediction": 1, "probabilities": [0.07383164763450623, 0.9220259785652161, 0.0041423519141972065], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-AR", "prediction": 1, "probabilities": [0.22971276938915253, 0.7596734762191772, 0.010613687336444855], "tokens": 58} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-DIFFUSION", "prediction": 1, "probabilities": [0.05981966108083725, 0.9343453645706177, 0.005834957119077444], "tokens": 52} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-BOTH", "prediction": 1, "probabilities": [0.17597955465316772, 0.8158330321311951, 0.008187456987798214], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-AR", "prediction": 1, "probabilities": [0.16877077519893646, 0.8112145066261292, 0.020014701411128044], "tokens": 59} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-DIFFUSION", "prediction": 1, "probabilities": [0.10481210052967072, 0.8791021108627319, 0.016085801646113396], "tokens": 55} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-BOTH", "prediction": 1, "probabilities": [0.08450927585363388, 0.8988321423530579, 0.016658615320920944], "tokens": 53} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-AR", "prediction": 1, "probabilities": [0.09692015498876572, 0.8916558027267456, 0.011424066498875618], "tokens": 42} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-DIFFUSION", "prediction": 1, "probabilities": [0.07338619977235794, 0.9192401766777039, 0.00737362913787365], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-BOTH", "prediction": 1, "probabilities": [0.09523030370473862, 0.8893879055976868, 0.015381810255348682], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-AR", "prediction": 1, "probabilities": [0.14335861802101135, 0.8041240572929382, 0.052517250180244446], "tokens": 42} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-DIFFUSION", "prediction": 1, "probabilities": [0.1721165031194687, 0.7945120334625244, 0.033371470868587494], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-BOTH", "prediction": 1, "probabilities": [0.09816297143697739, 0.879672110080719, 0.022164909169077873], "tokens": 37} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-AR", "prediction": 1, "probabilities": [0.11493905633687973, 0.8754770755767822, 0.009583847597241402], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-DIFFUSION", "prediction": 1, "probabilities": [0.08009988069534302, 0.9114881753921509, 0.008411929942667484], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-BOTH", "prediction": 1, "probabilities": [0.09357903897762299, 0.9010250568389893, 0.005395884159952402], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-AR", "prediction": 1, "probabilities": [0.05378357693552971, 0.9433642625808716, 0.0028522294014692307], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-DIFFUSION", "prediction": 1, "probabilities": [0.07451089471578598, 0.918745219707489, 0.006743822246789932], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-BOTH", "prediction": 1, "probabilities": [0.057666487991809845, 0.9370042085647583, 0.005329307168722153], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-AR", "prediction": 1, "probabilities": [0.16721968352794647, 0.8228096961975098, 0.009970596991479397], "tokens": 41} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-DIFFUSION", "prediction": 1, "probabilities": [0.18143132328987122, 0.811884343624115, 0.006684386171400547], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-BOTH", "prediction": 1, "probabilities": [0.18328997492790222, 0.8018242120742798, 0.014885859563946724], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-AR", "prediction": 1, "probabilities": [0.1743806004524231, 0.7789387702941895, 0.046680599451065063], "tokens": 43} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-DIFFUSION", "prediction": 1, "probabilities": [0.21569886803627014, 0.7591670751571655, 0.025134071707725525], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-BOTH", "prediction": 1, "probabilities": [0.08089827746152878, 0.886850893497467, 0.03225082531571388], "tokens": 39} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR", "prediction": 1, "probabilities": [0.22085128724575043, 0.6242893934249878, 0.15485930442810059], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-DIFFUSION", "prediction": 1, "probabilities": [0.07090184837579727, 0.8112279176712036, 0.11787019670009613], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-BOTH", "prediction": 1, "probabilities": [0.15724435448646545, 0.608342707157135, 0.23441292345523834], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-AR", "prediction": 2, "probabilities": [0.28455373644828796, 0.30788710713386536, 0.40755918622016907], "tokens": 32} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2756631374359131, 0.5512720346450806, 0.17306488752365112], "tokens": 30} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-BOTH", "prediction": 2, "probabilities": [0.2712460458278656, 0.34070172905921936, 0.38805222511291504], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-AR", "prediction": 0, "probabilities": [0.5774213075637817, 0.3299756944179535, 0.09260297566652298], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-DIFFUSION", "prediction": 1, "probabilities": [0.29148009419441223, 0.5711893439292908, 0.13733059167861938], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-BOTH", "prediction": 1, "probabilities": [0.2651045620441437, 0.45505398511886597, 0.27984148263931274], "tokens": 35} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-AR", "prediction": 0, "probabilities": [0.4717796742916107, 0.45456764101982117, 0.07365269958972931], "tokens": 39} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-DIFFUSION", "prediction": 1, "probabilities": [0.34043648838996887, 0.5474264621734619, 0.11213706433773041], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-BOTH", "prediction": 0, "probabilities": [0.46327129006385803, 0.38741227984428406, 0.1493164598941803], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-AR", "prediction": 0, "probabilities": [0.7988273501396179, 0.09849536418914795, 0.10267730802297592], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-DIFFUSION", "prediction": 1, "probabilities": [0.2598639130592346, 0.6184419989585876, 0.12169407308101654], "tokens": 32} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-BOTH", "prediction": 0, "probabilities": [0.8880519270896912, 0.04479566216468811, 0.06715243309736252], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-AR", "prediction": 0, "probabilities": [0.4210033118724823, 0.19662261009216309, 0.3823740482330322], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-DIFFUSION", "prediction": 0, "probabilities": [0.4803394675254822, 0.26779091358184814, 0.2518695890903473], "tokens": 34} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-BOTH", "prediction": 2, "probabilities": [0.35161611437797546, 0.2582070827484131, 0.3901767432689667], "tokens": 32} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-AR", "prediction": 1, "probabilities": [0.04640871286392212, 0.9436850547790527, 0.009906234219670296], "tokens": 85} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-DIFFUSION", "prediction": 1, "probabilities": [0.02836010605096817, 0.9645870923995972, 0.0070527466014027596], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-BOTH", "prediction": 1, "probabilities": [0.036728329956531525, 0.9566148519515991, 0.006656809244304895], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-AR", "prediction": 1, "probabilities": [0.033623263239860535, 0.9484388828277588, 0.017937811091542244], "tokens": 84} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0254281684756279, 0.9636054039001465, 0.010966463014483452], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-BOTH", "prediction": 1, "probabilities": [0.029113268479704857, 0.9607134461402893, 0.010173220187425613], "tokens": 79} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-AR", "prediction": 1, "probabilities": [0.041382044553756714, 0.9502146244049072, 0.008403293788433075], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-DIFFUSION", "prediction": 1, "probabilities": [0.04367027431726456, 0.9482556581497192, 0.008074101060628891], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-BOTH", "prediction": 1, "probabilities": [0.046660225838422775, 0.9471609592437744, 0.006178836803883314], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-AR", "prediction": 1, "probabilities": [0.02956661768257618, 0.9657600522041321, 0.004673371557146311], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-DIFFUSION", "prediction": 1, "probabilities": [0.04433892294764519, 0.949070155620575, 0.006590896751731634], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-BOTH", "prediction": 1, "probabilities": [0.030687352642416954, 0.964823305606842, 0.0044893804006278515], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-AR", "prediction": 1, "probabilities": [0.06764428317546844, 0.9233530759811401, 0.009002641774713993], "tokens": 84} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-DIFFUSION", "prediction": 1, "probabilities": [0.06025182828307152, 0.9335893988609314, 0.0061587379314005375], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-BOTH", "prediction": 1, "probabilities": [0.06721283495426178, 0.9256169199943542, 0.007170281372964382], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-AR", "prediction": 1, "probabilities": [0.041919123381376266, 0.9414636492729187, 0.016617225483059883], "tokens": 85} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-DIFFUSION", "prediction": 1, "probabilities": [0.045897629112005234, 0.9432991147041321, 0.010803287848830223], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-BOTH", "prediction": 1, "probabilities": [0.026482680812478065, 0.9626238942146301, 0.010893460363149643], "tokens": 81} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-AR", "prediction": 1, "probabilities": [0.044949937611818314, 0.9466550946235657, 0.00839498545974493], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-DIFFUSION", "prediction": 1, "probabilities": [0.028589770197868347, 0.96563720703125, 0.005773062352091074], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-BOTH", "prediction": 1, "probabilities": [0.04932140186429024, 0.9398764371871948, 0.010802086442708969], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-AR", "prediction": 1, "probabilities": [0.13318303227424622, 0.8371590375900269, 0.02965794876217842], "tokens": 39} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-DIFFUSION", "prediction": 1, "probabilities": [0.10816443711519241, 0.8780742883682251, 0.013761306181550026], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-BOTH", "prediction": 1, "probabilities": [0.12450145930051804, 0.8456953763961792, 0.029803207144141197], "tokens": 38} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-AR", "prediction": 1, "probabilities": [0.16644825041294098, 0.8260818123817444, 0.007469973061233759], "tokens": 44} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-DIFFUSION", "prediction": 1, "probabilities": [0.13656720519065857, 0.8575998544692993, 0.005832979921251535], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-BOTH", "prediction": 1, "probabilities": [0.12057139724493027, 0.8623290657997131, 0.01709948666393757], "tokens": 41} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-AR", "prediction": 1, "probabilities": [0.1377178281545639, 0.8565324544906616, 0.005749685689806938], "tokens": 45} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1219349354505539, 0.8714370131492615, 0.006628070026636124], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-BOTH", "prediction": 1, "probabilities": [0.17654170095920563, 0.8154636025428772, 0.00799472164362669], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-AR", "prediction": 1, "probabilities": [0.37454545497894287, 0.6138697266578674, 0.01158489752560854], "tokens": 44} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-DIFFUSION", "prediction": 1, "probabilities": [0.09382703900337219, 0.896758496761322, 0.009414521045982838], "tokens": 38} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-BOTH", "prediction": 0, "probabilities": [0.5342884659767151, 0.4546276032924652, 0.011083981953561306], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-AR", "prediction": 1, "probabilities": [0.25700220465660095, 0.7000887989997864, 0.04290901497006416], "tokens": 45} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-DIFFUSION", "prediction": 1, "probabilities": [0.3043771982192993, 0.6606468558311462, 0.034975938498973846], "tokens": 41} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-BOTH", "prediction": 1, "probabilities": [0.2033139020204544, 0.7523494958877563, 0.044336624443531036], "tokens": 39} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-AR", "prediction": 1, "probabilities": [0.07375635951757431, 0.915156364440918, 0.011087250895798206], "tokens": 36} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-DIFFUSION", "prediction": 1, "probabilities": [0.0988611951470375, 0.891201913356781, 0.009936965070664883], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-BOTH", "prediction": 1, "probabilities": [0.07162721455097198, 0.9163768291473389, 0.01199598703533411], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-AR", "prediction": 1, "probabilities": [0.1369812935590744, 0.7648305296897888, 0.09818816184997559], "tokens": 35} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2372530996799469, 0.7008138298988342, 0.06193307414650917], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-BOTH", "prediction": 1, "probabilities": [0.06623252481222153, 0.9108762145042419, 0.022891268134117126], "tokens": 30} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-AR", "prediction": 1, "probabilities": [0.16184081137180328, 0.8261433243751526, 0.012015929445624352], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-DIFFUSION", "prediction": 1, "probabilities": [0.1001565158367157, 0.8890600800514221, 0.010783465579152107], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-BOTH", "prediction": 1, "probabilities": [0.11350634694099426, 0.8792627453804016, 0.007230913732200861], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-AR", "prediction": 1, "probabilities": [0.05365598946809769, 0.9441065788269043, 0.0022373634856194258], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-DIFFUSION", "prediction": 1, "probabilities": [0.09121338278055191, 0.9027721285820007, 0.006014480721205473], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-BOTH", "prediction": 1, "probabilities": [0.05714982748031616, 0.9379231929779053, 0.004926918540149927], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-AR", "prediction": 1, "probabilities": [0.23215417563915253, 0.7553923726081848, 0.012453481554985046], "tokens": 35} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-DIFFUSION", "prediction": 1, "probabilities": [0.2389174997806549, 0.7552087306976318, 0.005873734597116709], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-BOTH", "prediction": 1, "probabilities": [0.27557897567749023, 0.7086207270622253, 0.01580035500228405], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-AR", "prediction": 1, "probabilities": [0.21078692376613617, 0.7273012399673462, 0.06191185861825943], "tokens": 36} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-DIFFUSION", "prediction": 1, "probabilities": [0.22393716871738434, 0.7416506409645081, 0.03441215679049492], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-BOTH", "prediction": 1, "probabilities": [0.04566545784473419, 0.9184287786483765, 0.035905733704566956], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-AR", "prediction": 1, "probabilities": [0.06844019144773483, 0.912346601486206, 0.019213279709219933], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-DIFFUSION", "prediction": 1, "probabilities": [0.04304074868559837, 0.9454485177993774, 0.011510781943798065], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-BOTH", "prediction": 1, "probabilities": [0.06084101274609566, 0.9189663529396057, 0.020192613825201988], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-AR", "prediction": 1, "probabilities": [0.20665034651756287, 0.7154941558837891, 0.07785553485155106], "tokens": 29} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-DIFFUSION", "prediction": 1, "probabilities": [0.26724621653556824, 0.6882490515708923, 0.04450474679470062], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-BOTH", "prediction": 1, "probabilities": [0.19605980813503265, 0.7397854924201965, 0.0641547292470932], "tokens": 28} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-AR", "prediction": 1, "probabilities": [0.35395652055740356, 0.6366548538208008, 0.009388661943376064], "tokens": 34} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-DIFFUSION", "prediction": 1, "probabilities": [0.1648223102092743, 0.8265346884727478, 0.008643025532364845], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-BOTH", "prediction": 1, "probabilities": [0.21491862833499908, 0.7605993747711182, 0.02448197640478611], "tokens": 31} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-AR", "prediction": 1, "probabilities": [0.24538515508174896, 0.7473013401031494, 0.0073134819976985455], "tokens": 35} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-DIFFUSION", "prediction": 1, "probabilities": [0.20063075423240662, 0.7907324433326721, 0.008636769838631153], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-BOTH", "prediction": 1, "probabilities": [0.3197389245033264, 0.6713376641273499, 0.00892338715493679], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-AR", "prediction": 0, "probabilities": [0.7245038747787476, 0.2571401596069336, 0.018355999141931534], "tokens": 34} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-DIFFUSION", "prediction": 1, "probabilities": [0.2441677451133728, 0.7422947883605957, 0.013537446968257427], "tokens": 28} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-BOTH", "prediction": 0, "probabilities": [0.7949944138526917, 0.1938064992427826, 0.011199025437235832], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-AR", "prediction": 1, "probabilities": [0.4112771153450012, 0.5092663764953613, 0.07945649325847626], "tokens": 35} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-DIFFUSION", "prediction": 1, "probabilities": [0.45446938276290894, 0.4835195541381836, 0.06201106682419777], "tokens": 31} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-BOTH", "prediction": 1, "probabilities": [0.36605212092399597, 0.5401951670646667, 0.0937526747584343], "tokens": 29} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-AR", "prediction": 1, "probabilities": [0.08469508588314056, 0.9067192077636719, 0.008585692383348942], "tokens": 47} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-DIFFUSION", "prediction": 1, "probabilities": [0.05767930671572685, 0.937267005443573, 0.00505374139174819], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-BOTH", "prediction": 1, "probabilities": [0.06347329914569855, 0.930488646030426, 0.0060380301438272], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-AR", "prediction": 1, "probabilities": [0.09387824684381485, 0.8733844757080078, 0.03273731470108032], "tokens": 46} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-DIFFUSION", "prediction": 1, "probabilities": [0.09849932044744492, 0.8800146579742432, 0.0214859526604414], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-BOTH", "prediction": 1, "probabilities": [0.0690765529870987, 0.9146188497543335, 0.016304554417729378], "tokens": 41} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-AR", "prediction": 1, "probabilities": [0.07674665004014969, 0.9175366163253784, 0.00571672897785902], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-DIFFUSION", "prediction": 1, "probabilities": [0.061751119792461395, 0.9329231381416321, 0.0053257206454873085], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-BOTH", "prediction": 1, "probabilities": [0.07520754635334015, 0.9206750988960266, 0.0041173528879880905], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-AR", "prediction": 1, "probabilities": [0.03531201556324959, 0.96305251121521, 0.0016355137340724468], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-DIFFUSION", "prediction": 1, "probabilities": [0.06236795708537102, 0.9341034889221191, 0.003528617788106203], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-BOTH", "prediction": 1, "probabilities": [0.050550948828458786, 0.9465883374214172, 0.002860745182260871], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-AR", "prediction": 1, "probabilities": [0.13903018832206726, 0.8538382053375244, 0.007131602149456739], "tokens": 46} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-DIFFUSION", "prediction": 1, "probabilities": [0.1200660690665245, 0.8759562969207764, 0.003977630287408829], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-BOTH", "prediction": 1, "probabilities": [0.1701645404100418, 0.8214656710624695, 0.008369749411940575], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-AR", "prediction": 1, "probabilities": [0.10688938200473785, 0.8693612217903137, 0.023749427869915962], "tokens": 47} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-DIFFUSION", "prediction": 1, "probabilities": [0.1452074497938156, 0.8406352996826172, 0.014157264493405819], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-BOTH", "prediction": 1, "probabilities": [0.04702829569578171, 0.9367815256118774, 0.01619010791182518], "tokens": 43} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-AR", "prediction": 1, "probabilities": [0.07562360167503357, 0.9010998606681824, 0.02327657863497734], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-DIFFUSION", "prediction": 1, "probabilities": [0.04925190284848213, 0.9414920806884766, 0.009256028570234776], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-BOTH", "prediction": 1, "probabilities": [0.10462121665477753, 0.8790321946144104, 0.016346534714102745], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-AR", "prediction": 1, "probabilities": [0.09338448196649551, 0.868768036365509, 0.03784747049212456], "tokens": 54} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07876785844564438, 0.8950842618942261, 0.026147857308387756], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-BOTH", "prediction": 1, "probabilities": [0.12543576955795288, 0.8338372707366943, 0.040726982057094574], "tokens": 53} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-AR", "prediction": 1, "probabilities": [0.13933613896369934, 0.844121515750885, 0.016542289406061172], "tokens": 59} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-DIFFUSION", "prediction": 1, "probabilities": [0.12900564074516296, 0.8605313897132874, 0.010462947189807892], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-BOTH", "prediction": 1, "probabilities": [0.17241308093070984, 0.8047758936882019, 0.0228110458701849], "tokens": 56} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-AR", "prediction": 1, "probabilities": [0.1368270069360733, 0.8525294065475464, 0.010643611662089825], "tokens": 60} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1086878702044487, 0.8811138868331909, 0.010198247618973255], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-BOTH", "prediction": 1, "probabilities": [0.2369394153356552, 0.7498982548713684, 0.013162335380911827], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-AR", "prediction": 1, "probabilities": [0.33315137028694153, 0.6480520367622375, 0.018796579912304878], "tokens": 59} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-DIFFUSION", "prediction": 1, "probabilities": [0.08632635325193405, 0.8998903036117554, 0.013783375732600689], "tokens": 53} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-BOTH", "prediction": 1, "probabilities": [0.3392239212989807, 0.6376457214355469, 0.023130426183342934], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-AR", "prediction": 1, "probabilities": [0.24054987728595734, 0.6886298656463623, 0.07082029432058334], "tokens": 60} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-DIFFUSION", "prediction": 1, "probabilities": [0.13945502042770386, 0.8140810132026672, 0.046464018523693085], "tokens": 56} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-BOTH", "prediction": 1, "probabilities": [0.11151716858148575, 0.8530034422874451, 0.03547937795519829], "tokens": 54} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-AR", "prediction": 1, "probabilities": [0.34097087383270264, 0.5068780183792114, 0.15215112268924713], "tokens": 33} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-DIFFUSION", "prediction": 1, "probabilities": [0.17968560755252838, 0.6699467301368713, 0.1503676176071167], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH", "prediction": 1, "probabilities": [0.3370288610458374, 0.3541170358657837, 0.3088541328907013], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-AR", "prediction": 2, "probabilities": [0.31841379404067993, 0.23972353339195251, 0.44186270236968994], "tokens": 32} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-DIFFUSION", "prediction": 2, "probabilities": [0.4075321853160858, 0.1739189326763153, 0.4185488224029541], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-BOTH", "prediction": 1, "probabilities": [0.2697162628173828, 0.4667767882347107, 0.2635069787502289], "tokens": 27} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-AR", "prediction": 0, "probabilities": [0.4949442446231842, 0.4177636504173279, 0.08729209750890732], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-DIFFUSION", "prediction": 0, "probabilities": [0.4773167073726654, 0.4255923330783844, 0.09709099680185318], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-BOTH", "prediction": 1, "probabilities": [0.44367727637290955, 0.49143168330192566, 0.06489107012748718], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-AR", "prediction": 1, "probabilities": [0.33775779604911804, 0.6273531913757324, 0.03488900139927864], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-DIFFUSION", "prediction": 1, "probabilities": [0.427376925945282, 0.4687653183937073, 0.10385768860578537], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-BOTH", "prediction": 1, "probabilities": [0.2792438268661499, 0.6146544814109802, 0.1061016246676445], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-AR", "prediction": 0, "probabilities": [0.6457979083061218, 0.2687251567840576, 0.08547694236040115], "tokens": 32} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-DIFFUSION", "prediction": 0, "probabilities": [0.7832170724868774, 0.17365019023418427, 0.04313276335597038], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-BOTH", "prediction": 0, "probabilities": [0.727013111114502, 0.157637357711792, 0.11534947901964188], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-AR", "prediction": 0, "probabilities": [0.43661826848983765, 0.13658271729946136, 0.4267989993095398], "tokens": 33} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-DIFFUSION", "prediction": 0, "probabilities": [0.6365784406661987, 0.17172925174236298, 0.1916922777891159], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-BOTH", "prediction": 2, "probabilities": [0.12150456756353378, 0.1951274424791336, 0.6833680272102356], "tokens": 29} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-AR", "prediction": 1, "probabilities": [0.10980681329965591, 0.8559392690658569, 0.034253884106874466], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-DIFFUSION", "prediction": 1, "probabilities": [0.06726448982954025, 0.9097042083740234, 0.023031357675790787], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-BOTH", "prediction": 1, "probabilities": [0.15172460675239563, 0.7929716110229492, 0.05530376359820366], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-AR", "prediction": 1, "probabilities": [0.2729003429412842, 0.6070871949195862, 0.12001245468854904], "tokens": 29} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2550910413265228, 0.6609871983528137, 0.08392178267240524], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-BOTH", "prediction": 1, "probabilities": [0.2888050377368927, 0.5783653855323792, 0.13282953202724457], "tokens": 28} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-AR", "prediction": 0, "probabilities": [0.5543198585510254, 0.4297654926776886, 0.01591467298567295], "tokens": 34} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-DIFFUSION", "prediction": 1, "probabilities": [0.24409644305706024, 0.7373910546302795, 0.018512461334466934], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-BOTH", "prediction": 1, "probabilities": [0.35752055048942566, 0.5754354596138, 0.06704398989677429], "tokens": 31} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-AR", "prediction": 1, "probabilities": [0.4104100465774536, 0.5775038003921509, 0.012086200527846813], "tokens": 35} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-DIFFUSION", "prediction": 1, "probabilities": [0.3282315731048584, 0.647430419921875, 0.024338090792298317], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-BOTH", "prediction": 0, "probabilities": [0.5736560225486755, 0.40787234902381897, 0.018471650779247284], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-AR", "prediction": 0, "probabilities": [0.8647586703300476, 0.11274320632219315, 0.02249804325401783], "tokens": 34} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-DIFFUSION", "prediction": 1, "probabilities": [0.3071434199810028, 0.6620287895202637, 0.030827747657895088], "tokens": 28} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-BOTH", "prediction": 0, "probabilities": [0.936271071434021, 0.05038103088736534, 0.013347824104130268], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-AR", "prediction": 0, "probabilities": [0.6231698393821716, 0.26041874289512634, 0.11641138046979904], "tokens": 35} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-DIFFUSION", "prediction": 0, "probabilities": [0.4849383533000946, 0.3878086805343628, 0.127252995967865], "tokens": 31} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-BOTH", "prediction": 1, "probabilities": [0.38576850295066833, 0.44832316040992737, 0.1659083515405655], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-AR", "prediction": 1, "probabilities": [0.18351349234580994, 0.7795410752296448, 0.0369454100728035], "tokens": 30} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-DIFFUSION", "prediction": 1, "probabilities": [0.12766295671463013, 0.8452103734016418, 0.02712664008140564], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-BOTH", "prediction": 1, "probabilities": [0.1737106442451477, 0.7745549082756042, 0.05173446983098984], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-AR", "prediction": 1, "probabilities": [0.2819540798664093, 0.5305783152580261, 0.18746760487556458], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-DIFFUSION", "prediction": 0, "probabilities": [0.4742015302181244, 0.3872399628162384, 0.13855858147144318], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-BOTH", "prediction": 1, "probabilities": [0.22110046446323395, 0.7089929580688477, 0.06990658491849899], "tokens": 24} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-AR", "prediction": 1, "probabilities": [0.3479655086994171, 0.6310637593269348, 0.020970778539776802], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-DIFFUSION", "prediction": 1, "probabilities": [0.29834797978401184, 0.680467963218689, 0.021184062585234642], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-BOTH", "prediction": 1, "probabilities": [0.2768212556838989, 0.7108798027038574, 0.012298854067921638], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-AR", "prediction": 1, "probabilities": [0.14822280406951904, 0.8443175554275513, 0.007459686137735844], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-DIFFUSION", "prediction": 1, "probabilities": [0.2657630741596222, 0.7163896560668945, 0.017847226932644844], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-BOTH", "prediction": 1, "probabilities": [0.15985040366649628, 0.8256184458732605, 0.014531140215694904], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-AR", "prediction": 1, "probabilities": [0.454793244600296, 0.5159580111503601, 0.029248744249343872], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-DIFFUSION", "prediction": 1, "probabilities": [0.4792669117450714, 0.5071325302124023, 0.013600615784525871], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-BOTH", "prediction": 0, "probabilities": [0.49388110637664795, 0.4655664265155792, 0.04055243730545044], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-AR", "prediction": 0, "probabilities": [0.4486352205276489, 0.3707672953605652, 0.18059757351875305], "tokens": 30} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-DIFFUSION", "prediction": 0, "probabilities": [0.6227002739906311, 0.32093676924705505, 0.056362953037023544], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-BOTH", "prediction": 1, "probabilities": [0.17036482691764832, 0.6601830124855042, 0.16945217549800873], "tokens": 26} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-AR", "prediction": 1, "probabilities": [0.024081891402602196, 0.9716491103172302, 0.004268977325409651], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-DIFFUSION", "prediction": 1, "probabilities": [0.016231698915362358, 0.9816957712173462, 0.0020725869107991457], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-BOTH", "prediction": 1, "probabilities": [0.020620156079530716, 0.975439727306366, 0.003940084017813206], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-AR", "prediction": 1, "probabilities": [0.020273473113775253, 0.973095178604126, 0.006631307303905487], "tokens": 56} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-DIFFUSION", "prediction": 1, "probabilities": [0.022757278755307198, 0.9704795479774475, 0.006763157434761524], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-BOTH", "prediction": 1, "probabilities": [0.0304244477301836, 0.9613741040229797, 0.008201510645449162], "tokens": 55} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-AR", "prediction": 1, "probabilities": [0.04350036010146141, 0.9528993368148804, 0.003600325435400009], "tokens": 61} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-DIFFUSION", "prediction": 1, "probabilities": [0.029604313895106316, 0.9681974649429321, 0.002198267262428999], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-BOTH", "prediction": 1, "probabilities": [0.029958685860037804, 0.9662826657295227, 0.0037586786784231663], "tokens": 58} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-AR", "prediction": 1, "probabilities": [0.02870667725801468, 0.9693936109542847, 0.001899736002087593], "tokens": 62} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-DIFFUSION", "prediction": 1, "probabilities": [0.0247273501008749, 0.97257000207901, 0.002702603582292795], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-BOTH", "prediction": 1, "probabilities": [0.035762324929237366, 0.9612489342689514, 0.0029886867851018906], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-AR", "prediction": 1, "probabilities": [0.08129649609327316, 0.9139421582221985, 0.0047613210044801235], "tokens": 61} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-DIFFUSION", "prediction": 1, "probabilities": [0.02954583615064621, 0.967073380947113, 0.003380831331014633], "tokens": 55} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-BOTH", "prediction": 1, "probabilities": [0.060038674622774124, 0.9360052347183228, 0.0039561353623867035], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-AR", "prediction": 1, "probabilities": [0.060972247272729874, 0.9271680116653442, 0.011859782971441746], "tokens": 62} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-DIFFUSION", "prediction": 1, "probabilities": [0.04097817838191986, 0.9486884474754333, 0.01033330149948597], "tokens": 58} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-BOTH", "prediction": 1, "probabilities": [0.029310546815395355, 0.9624460935592651, 0.008243337273597717], "tokens": 56} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-AR", "prediction": 1, "probabilities": [0.2686540186405182, 0.653056263923645, 0.07828978449106216], "tokens": 28} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-DIFFUSION", "prediction": 1, "probabilities": [0.1900860220193863, 0.7480887770652771, 0.06182517111301422], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-BOTH", "prediction": 1, "probabilities": [0.28136542439460754, 0.5006031394004822, 0.21803142130374908], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-AR", "prediction": 2, "probabilities": [0.31722408533096313, 0.2598035931587219, 0.4229723811149597], "tokens": 27} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-DIFFUSION", "prediction": 0, "probabilities": [0.45004189014434814, 0.15371602773666382, 0.3962421119213104], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-BOTH", "prediction": 1, "probabilities": [0.2627190351486206, 0.567561686038971, 0.16971924901008606], "tokens": 22} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-AR", "prediction": 0, "probabilities": [0.5486798882484436, 0.41431403160095215, 0.037006039172410965], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-DIFFUSION", "prediction": 0, "probabilities": [0.5371905565261841, 0.3955394923686981, 0.0672699362039566], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-BOTH", "prediction": 1, "probabilities": [0.3261764347553253, 0.6498467326164246, 0.023976769298315048], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-AR", "prediction": 1, "probabilities": [0.226170614361763, 0.7596568465232849, 0.014172476716339588], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-DIFFUSION", "prediction": 1, "probabilities": [0.4092291593551636, 0.5430411696434021, 0.04772967845201492], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-BOTH", "prediction": 1, "probabilities": [0.22938871383666992, 0.7420359253883362, 0.028575314208865166], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-AR", "prediction": 0, "probabilities": [0.7802357077598572, 0.1970731019973755, 0.022691236808896065], "tokens": 27} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-DIFFUSION", "prediction": 0, "probabilities": [0.8046485781669617, 0.17755627632141113, 0.017795177176594734], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-BOTH", "prediction": 0, "probabilities": [0.7638722062110901, 0.17275679111480713, 0.06337100267410278], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-AR", "prediction": 0, "probabilities": [0.6356345415115356, 0.10941381752490997, 0.25495168566703796], "tokens": 28} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-DIFFUSION", "prediction": 0, "probabilities": [0.7377748489379883, 0.16630378365516663, 0.09592139720916748], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-BOTH", "prediction": 2, "probabilities": [0.18844053149223328, 0.33097192645072937, 0.48058751225471497], "tokens": 24} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-AR", "prediction": 1, "probabilities": [0.10873182862997055, 0.8690492510795593, 0.0222189798951149], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-DIFFUSION", "prediction": 1, "probabilities": [0.062063902616500854, 0.9231917858123779, 0.014744340442121029], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-BOTH", "prediction": 1, "probabilities": [0.12091376632452011, 0.849014401435852, 0.03007185459136963], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-AR", "prediction": 1, "probabilities": [0.3838660717010498, 0.5368819236755371, 0.07925206422805786], "tokens": 27} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2989599406719208, 0.6383981704711914, 0.06264190375804901], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-BOTH", "prediction": 1, "probabilities": [0.3355294167995453, 0.5529895424842834, 0.11148109287023544], "tokens": 26} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-AR", "prediction": 0, "probabilities": [0.5198369026184082, 0.4703652262687683, 0.009797943755984306], "tokens": 32} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-DIFFUSION", "prediction": 1, "probabilities": [0.35232967138290405, 0.6358532905578613, 0.011816992424428463], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-BOTH", "prediction": 1, "probabilities": [0.3423967659473419, 0.6107016801834106, 0.04690151661634445], "tokens": 29} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-AR", "prediction": 1, "probabilities": [0.41176292300224304, 0.5808355808258057, 0.007401501294225454], "tokens": 33} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-DIFFUSION", "prediction": 1, "probabilities": [0.4516637921333313, 0.5325706601142883, 0.015765558928251266], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-BOTH", "prediction": 0, "probabilities": [0.5928969383239746, 0.392479807138443, 0.014623329974710941], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-AR", "prediction": 0, "probabilities": [0.8386108875274658, 0.1421908736228943, 0.01919817551970482], "tokens": 32} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-DIFFUSION", "prediction": 1, "probabilities": [0.2709364891052246, 0.7080004811286926, 0.02106305956840515], "tokens": 26} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-BOTH", "prediction": 0, "probabilities": [0.9134799838066101, 0.073137566447258, 0.013382509350776672], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-AR", "prediction": 0, "probabilities": [0.6586618423461914, 0.24671660363674164, 0.09462153911590576], "tokens": 33} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-DIFFUSION", "prediction": 0, "probabilities": [0.639703094959259, 0.2729050815105438, 0.08739180862903595], "tokens": 29} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-BOTH", "prediction": 0, "probabilities": [0.4324253797531128, 0.38040071725845337, 0.18717381358146667], "tokens": 27} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-AR", "prediction": 1, "probabilities": [0.07452661544084549, 0.8949946761131287, 0.0304787028580904], "tokens": 37} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-DIFFUSION", "prediction": 1, "probabilities": [0.05176080763339996, 0.9312624335289001, 0.01697673462331295], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-BOTH", "prediction": 1, "probabilities": [0.03578871116042137, 0.9388234615325928, 0.025387883186340332], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-AR", "prediction": 1, "probabilities": [0.08221518993377686, 0.7783441543579102, 0.13944068551063538], "tokens": 37} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-DIFFUSION", "prediction": 1, "probabilities": [0.10826187580823898, 0.7762597799301147, 0.1154782772064209], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-BOTH", "prediction": 1, "probabilities": [0.04003174975514412, 0.9174095988273621, 0.042558703571558], "tokens": 32} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-AR", "prediction": 1, "probabilities": [0.0759299248456955, 0.9067480564117432, 0.017322035506367683], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-DIFFUSION", "prediction": 1, "probabilities": [0.09411165863275528, 0.8860394954681396, 0.01984887756407261], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-BOTH", "prediction": 1, "probabilities": [0.08169753849506378, 0.9061571955680847, 0.012145234271883965], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-AR", "prediction": 1, "probabilities": [0.018546871840953827, 0.9777202606201172, 0.003732840297743678], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-DIFFUSION", "prediction": 1, "probabilities": [0.06057378649711609, 0.92849200963974, 0.010934202000498772], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-BOTH", "prediction": 1, "probabilities": [0.02696922980248928, 0.9648662805557251, 0.008164520375430584], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-AR", "prediction": 1, "probabilities": [0.14886441826820374, 0.8264259696006775, 0.02470957301557064], "tokens": 36} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-DIFFUSION", "prediction": 1, "probabilities": [0.16037523746490479, 0.8274989724159241, 0.012125729583203793], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-BOTH", "prediction": 1, "probabilities": [0.20660817623138428, 0.7624478340148926, 0.030944013968110085], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-AR", "prediction": 1, "probabilities": [0.1152166947722435, 0.7807571887969971, 0.1040261760354042], "tokens": 38} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-DIFFUSION", "prediction": 1, "probabilities": [0.1626340001821518, 0.7897020578384399, 0.047663941979408264], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-BOTH", "prediction": 1, "probabilities": [0.03956203535199165, 0.8873537182807922, 0.073084257543087], "tokens": 34} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-AR", "prediction": 1, "probabilities": [0.05074549466371536, 0.9341455101966858, 0.015108980238437653], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-DIFFUSION", "prediction": 1, "probabilities": [0.024905703961849213, 0.969131588935852, 0.005962699186056852], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-BOTH", "prediction": 1, "probabilities": [0.03977743163704872, 0.9433389902114868, 0.016883570700883865], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-AR", "prediction": 1, "probabilities": [0.06544922292232513, 0.9060242772102356, 0.028526440262794495], "tokens": 53} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07783596217632294, 0.8981359004974365, 0.024028198793530464], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-BOTH", "prediction": 1, "probabilities": [0.08195184916257858, 0.8931254148483276, 0.024922722950577736], "tokens": 52} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-AR", "prediction": 1, "probabilities": [0.08799798786640167, 0.8995435833930969, 0.012458482757210732], "tokens": 58} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-DIFFUSION", "prediction": 1, "probabilities": [0.07025104761123657, 0.922463059425354, 0.0072858515195548534], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-BOTH", "prediction": 1, "probabilities": [0.06743193417787552, 0.9153427481651306, 0.017225347459316254], "tokens": 55} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-AR", "prediction": 1, "probabilities": [0.08175630122423172, 0.9083739519119263, 0.009869727306067944], "tokens": 59} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-DIFFUSION", "prediction": 1, "probabilities": [0.07503743469715118, 0.9177435040473938, 0.007219068706035614], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-BOTH", "prediction": 1, "probabilities": [0.10404834151268005, 0.8840416669845581, 0.01190999150276184], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-AR", "prediction": 1, "probabilities": [0.29205888509750366, 0.68916916847229, 0.018771938979625702], "tokens": 58} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-DIFFUSION", "prediction": 1, "probabilities": [0.09104283154010773, 0.897092878818512, 0.011864295229315758], "tokens": 52} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-BOTH", "prediction": 1, "probabilities": [0.31258776783943176, 0.6706022620201111, 0.016809996217489243], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-AR", "prediction": 1, "probabilities": [0.21632561087608337, 0.7211126685142517, 0.06256172060966492], "tokens": 59} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-DIFFUSION", "prediction": 1, "probabilities": [0.13099552690982819, 0.8166276812553406, 0.052376773208379745], "tokens": 55} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-BOTH", "prediction": 1, "probabilities": [0.08317887783050537, 0.8867194652557373, 0.030101720243692398], "tokens": 53} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-AR", "prediction": 1, "probabilities": [0.1616770178079605, 0.8146467208862305, 0.023676209151744843], "tokens": 30} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-DIFFUSION", "prediction": 1, "probabilities": [0.15888628363609314, 0.8234332203865051, 0.017680499702692032], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-BOTH", "prediction": 1, "probabilities": [0.193056121468544, 0.7588624954223633, 0.04808136448264122], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-AR", "prediction": 1, "probabilities": [0.28713223338127136, 0.5532437562942505, 0.15962402522563934], "tokens": 29} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-DIFFUSION", "prediction": 0, "probabilities": [0.4869982898235321, 0.4132409989833832, 0.09976071119308472], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-BOTH", "prediction": 1, "probabilities": [0.16377398371696472, 0.7634645700454712, 0.07276145368814468], "tokens": 24} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-AR", "prediction": 1, "probabilities": [0.30656397342681885, 0.6779853105545044, 0.015450656414031982], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-DIFFUSION", "prediction": 1, "probabilities": [0.21643325686454773, 0.7710565328598022, 0.012510214000940323], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-BOTH", "prediction": 1, "probabilities": [0.2317425161600113, 0.7584593296051025, 0.009798184968531132], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-AR", "prediction": 1, "probabilities": [0.12476567178964615, 0.8722321391105652, 0.0030022007413208485], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-DIFFUSION", "prediction": 1, "probabilities": [0.2025797963142395, 0.7878681421279907, 0.00955205224454403], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-BOTH", "prediction": 1, "probabilities": [0.19771194458007812, 0.79354327917099, 0.008744794875383377], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-AR", "prediction": 1, "probabilities": [0.3923453390598297, 0.5915049314498901, 0.016149764880537987], "tokens": 29} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-DIFFUSION", "prediction": 1, "probabilities": [0.49177461862564087, 0.5015097856521606, 0.0067156096920371056], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-BOTH", "prediction": 0, "probabilities": [0.6246941685676575, 0.35245615243911743, 0.022849712520837784], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-AR", "prediction": 0, "probabilities": [0.43869027495384216, 0.42541947960853577, 0.13589026033878326], "tokens": 30} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-DIFFUSION", "prediction": 0, "probabilities": [0.604336678981781, 0.34595945477485657, 0.04970387741923332], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-BOTH", "prediction": 1, "probabilities": [0.1377543807029724, 0.7633644342422485, 0.09888116270303726], "tokens": 26} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-AR", "prediction": 1, "probabilities": [0.12225434184074402, 0.8514963984489441, 0.02624923549592495], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-DIFFUSION", "prediction": 1, "probabilities": [0.07518506795167923, 0.9048416018486023, 0.019973326474428177], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-BOTH", "prediction": 1, "probabilities": [0.09536931663751602, 0.8788148760795593, 0.025815829634666443], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-AR", "prediction": 1, "probabilities": [0.14528006315231323, 0.810012698173523, 0.04470723867416382], "tokens": 55} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-DIFFUSION", "prediction": 1, "probabilities": [0.10847961157560349, 0.8499864339828491, 0.04153390973806381], "tokens": 53} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-BOTH", "prediction": 1, "probabilities": [0.11057388037443161, 0.8552589416503906, 0.03416716307401657], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-AR", "prediction": 1, "probabilities": [0.20181047916412354, 0.7712567448616028, 0.02693273313343525], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-DIFFUSION", "prediction": 1, "probabilities": [0.10798843950033188, 0.8751899600028992, 0.016821609809994698], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-BOTH", "prediction": 1, "probabilities": [0.12008063495159149, 0.8533457517623901, 0.026573603972792625], "tokens": 58} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-AR", "prediction": 1, "probabilities": [0.18646839261054993, 0.79871666431427, 0.01481489185243845], "tokens": 62} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-DIFFUSION", "prediction": 1, "probabilities": [0.14669696986675262, 0.8345139026641846, 0.018789101392030716], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-BOTH", "prediction": 1, "probabilities": [0.15099400281906128, 0.826366126537323, 0.02263985015451908], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-AR", "prediction": 1, "probabilities": [0.37764593958854675, 0.5960991978645325, 0.026254834607243538], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-DIFFUSION", "prediction": 1, "probabilities": [0.12169910967350006, 0.8528448939323425, 0.025455975905060768], "tokens": 55} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-BOTH", "prediction": 1, "probabilities": [0.4554417133331299, 0.5194826126098633, 0.025075729936361313], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-AR", "prediction": 1, "probabilities": [0.23203334212303162, 0.7105311155319214, 0.05743550509214401], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-DIFFUSION", "prediction": 1, "probabilities": [0.18483249843120575, 0.7598593831062317, 0.055308129638433456], "tokens": 57} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-BOTH", "prediction": 1, "probabilities": [0.14621540904045105, 0.8085188865661621, 0.045265741646289825], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-AR", "prediction": 1, "probabilities": [0.2989349663257599, 0.5497279763221741, 0.15133704245090485], "tokens": 59} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-DIFFUSION", "prediction": 1, "probabilities": [0.2282501757144928, 0.62653648853302, 0.14521335065364838], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-BOTH", "prediction": 1, "probabilities": [0.20011718571186066, 0.5743717551231384, 0.22551105916500092], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-AR", "prediction": 2, "probabilities": [0.2905658483505249, 0.35004448890686035, 0.35938963294029236], "tokens": 58} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-DIFFUSION", "prediction": 1, "probabilities": [0.30052807927131653, 0.3921654522418976, 0.3073064982891083], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-BOTH", "prediction": 1, "probabilities": [0.2669799327850342, 0.4600319564342499, 0.27298814058303833], "tokens": 53} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-AR", "prediction": 1, "probabilities": [0.3531420826911926, 0.5143290758132935, 0.13252884149551392], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-DIFFUSION", "prediction": 1, "probabilities": [0.30084383487701416, 0.5416712760925293, 0.15748488903045654], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-BOTH", "prediction": 1, "probabilities": [0.25438112020492554, 0.6328328251838684, 0.11278600990772247], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-AR", "prediction": 1, "probabilities": [0.25456783175468445, 0.6755349636077881, 0.06989724934101105], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-DIFFUSION", "prediction": 1, "probabilities": [0.29630473256111145, 0.5813713073730469, 0.12232393771409988], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-BOTH", "prediction": 1, "probabilities": [0.2188570499420166, 0.6765832304954529, 0.10455965995788574], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-AR", "prediction": 1, "probabilities": [0.3997581899166107, 0.46479281783103943, 0.13544894754886627], "tokens": 58} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-DIFFUSION", "prediction": 0, "probabilities": [0.5817944407463074, 0.31347519159317017, 0.10473034530878067], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-BOTH", "prediction": 0, "probabilities": [0.44231316447257996, 0.38342854380607605, 0.17425833642482758], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-AR", "prediction": 1, "probabilities": [0.3161203861236572, 0.35111817717552185, 0.33276140689849854], "tokens": 59} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-DIFFUSION", "prediction": 1, "probabilities": [0.39481353759765625, 0.39814087748527527, 0.2070455700159073], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-BOTH", "prediction": 1, "probabilities": [0.17945586144924164, 0.48010143637657166, 0.3404427170753479], "tokens": 55} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-AR", "prediction": 1, "probabilities": [0.00970204547047615, 0.9864440560340881, 0.003853833768516779], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-DIFFUSION", "prediction": 1, "probabilities": [0.005258819088339806, 0.993032693862915, 0.0017084799474105239], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-BOTH", "prediction": 1, "probabilities": [0.010253623127937317, 0.9865224957466125, 0.003223886713385582], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-AR", "prediction": 1, "probabilities": [0.014191058464348316, 0.978617787361145, 0.007191204931586981], "tokens": 62} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-DIFFUSION", "prediction": 1, "probabilities": [0.013945870101451874, 0.981999933719635, 0.0040541235357522964], "tokens": 60} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-BOTH", "prediction": 1, "probabilities": [0.01291092298924923, 0.9804593324661255, 0.006629799492657185], "tokens": 61} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-AR", "prediction": 1, "probabilities": [0.01944613829255104, 0.9774655103683472, 0.0030883820727467537], "tokens": 66} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-DIFFUSION", "prediction": 1, "probabilities": [0.010785052552819252, 0.9873334169387817, 0.0018814622890204191], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-BOTH", "prediction": 1, "probabilities": [0.01624160259962082, 0.9806726574897766, 0.003085789270699024], "tokens": 63} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-AR", "prediction": 1, "probabilities": [0.018952272832393646, 0.978665292263031, 0.002382406033575535], "tokens": 67} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-DIFFUSION", "prediction": 1, "probabilities": [0.011068307794630527, 0.986858606338501, 0.0020731317345052958], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-BOTH", "prediction": 1, "probabilities": [0.028357932344079018, 0.9694655537605286, 0.0021765625569969416], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-AR", "prediction": 1, "probabilities": [0.048990264534950256, 0.9464474320411682, 0.0045622568577528], "tokens": 66} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-DIFFUSION", "prediction": 1, "probabilities": [0.013964732177555561, 0.983125627040863, 0.00290959095582366], "tokens": 60} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-BOTH", "prediction": 1, "probabilities": [0.03810917213559151, 0.9585612416267395, 0.0033295834437012672], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-AR", "prediction": 1, "probabilities": [0.035270024091005325, 0.9557408094406128, 0.008989165537059307], "tokens": 68} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-DIFFUSION", "prediction": 1, "probabilities": [0.024062994867563248, 0.9681549072265625, 0.007782096043229103], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-BOTH", "prediction": 1, "probabilities": [0.01778969168663025, 0.9730314016342163, 0.009178940206766129], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-AR", "prediction": 1, "probabilities": [0.011114696972072124, 0.980805516242981, 0.008079782128334045], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-DIFFUSION", "prediction": 1, "probabilities": [0.008832544088363647, 0.9857429265975952, 0.005424531176686287], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-BOTH", "prediction": 1, "probabilities": [0.00728083960711956, 0.9866132140159607, 0.006105897948145866], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-AR", "prediction": 1, "probabilities": [0.009365681558847427, 0.9747809171676636, 0.01585337705910206], "tokens": 61} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-DIFFUSION", "prediction": 1, "probabilities": [0.007644513621926308, 0.9808526635169983, 0.011502866633236408], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-BOTH", "prediction": 1, "probabilities": [0.00818308349698782, 0.981332540512085, 0.010484345257282257], "tokens": 56} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-AR", "prediction": 1, "probabilities": [0.006869846023619175, 0.987725555896759, 0.005404637195169926], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-DIFFUSION", "prediction": 1, "probabilities": [0.009099933318793774, 0.9840983748435974, 0.006801688112318516], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-BOTH", "prediction": 1, "probabilities": [0.008963125757873058, 0.9859937429428101, 0.00504307122901082], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-AR", "prediction": 1, "probabilities": [0.004480691626667976, 0.993411123752594, 0.002108148764818907], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-DIFFUSION", "prediction": 1, "probabilities": [0.008201013319194317, 0.9873173832893372, 0.00448160246014595], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-BOTH", "prediction": 1, "probabilities": [0.005959948990494013, 0.9912188649177551, 0.0028212242759764194], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-AR", "prediction": 1, "probabilities": [0.016193697229027748, 0.9770697355270386, 0.006736539304256439], "tokens": 61} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-DIFFUSION", "prediction": 1, "probabilities": [0.01164405420422554, 0.9837816953659058, 0.004574241582304239], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-BOTH", "prediction": 1, "probabilities": [0.015338161028921604, 0.9778609871864319, 0.006800889037549496], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-AR", "prediction": 1, "probabilities": [0.008888277225196362, 0.9772845506668091, 0.013827142305672169], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-DIFFUSION", "prediction": 1, "probabilities": [0.0118118766695261, 0.9791069030761719, 0.009081193245947361], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-BOTH", "prediction": 1, "probabilities": [0.005962344817817211, 0.9833902716636658, 0.010647357441484928], "tokens": 58} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-AR", "prediction": 1, "probabilities": [0.07246603816747665, 0.8835915923118591, 0.04394236207008362], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-DIFFUSION", "prediction": 1, "probabilities": [0.020478785037994385, 0.9616071581840515, 0.017913997173309326], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-BOTH", "prediction": 1, "probabilities": [0.07119784504175186, 0.8829761147499084, 0.04582603648304939], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-AR", "prediction": 1, "probabilities": [0.22885161638259888, 0.6741450428962708, 0.09700337797403336], "tokens": 25} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-DIFFUSION", "prediction": 1, "probabilities": [0.163461834192276, 0.7922693490982056, 0.04426880553364754], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-BOTH", "prediction": 1, "probabilities": [0.14225536584854126, 0.7786113619804382, 0.07913323491811752], "tokens": 24} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-AR", "prediction": 1, "probabilities": [0.40769878029823303, 0.5814095139503479, 0.010891670361161232], "tokens": 30} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-DIFFUSION", "prediction": 1, "probabilities": [0.07852630317211151, 0.9117974042892456, 0.009676262736320496], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-BOTH", "prediction": 1, "probabilities": [0.16845498979091644, 0.771859347820282, 0.05968571826815605], "tokens": 27} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-AR", "prediction": 1, "probabilities": [0.18443503975868225, 0.8050287961959839, 0.010536106303334236], "tokens": 31} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-DIFFUSION", "prediction": 1, "probabilities": [0.12388987094163895, 0.8604207038879395, 0.015689408406615257], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-BOTH", "prediction": 1, "probabilities": [0.27144289016723633, 0.6986292600631714, 0.029927821829915047], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-AR", "prediction": 0, "probabilities": [0.849200427532196, 0.13242098689079285, 0.018378596752882004], "tokens": 30} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-DIFFUSION", "prediction": 1, "probabilities": [0.12382780760526657, 0.8605202436447144, 0.015651926398277283], "tokens": 24} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-BOTH", "prediction": 0, "probabilities": [0.9276393055915833, 0.06192617118358612, 0.010434543713927269], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-AR", "prediction": 0, "probabilities": [0.5643898844718933, 0.31292101740837097, 0.12268907576799393], "tokens": 31} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-DIFFUSION", "prediction": 1, "probabilities": [0.360359251499176, 0.5433877110481262, 0.09625304490327835], "tokens": 27} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-BOTH", "prediction": 1, "probabilities": [0.31407347321510315, 0.5376612544059753, 0.14826524257659912], "tokens": 25} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-AR", "prediction": 1, "probabilities": [0.07758182287216187, 0.9039220213890076, 0.018496178090572357], "tokens": 55} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-DIFFUSION", "prediction": 1, "probabilities": [0.0560014471411705, 0.9307242035865784, 0.013274396769702435], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-BOTH", "prediction": 1, "probabilities": [0.08477136492729187, 0.8932979106903076, 0.02193075232207775], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-AR", "prediction": 1, "probabilities": [0.09572293609380722, 0.8361082673072815, 0.06816878914833069], "tokens": 54} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-DIFFUSION", "prediction": 1, "probabilities": [0.08383214473724365, 0.8814928531646729, 0.03467496111989021], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-BOTH", "prediction": 1, "probabilities": [0.059228360652923584, 0.9197432994842529, 0.021028360351920128], "tokens": 49} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-AR", "prediction": 1, "probabilities": [0.1047663614153862, 0.8788799047470093, 0.01635371334850788], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-DIFFUSION", "prediction": 1, "probabilities": [0.08456791937351227, 0.900464653968811, 0.014967428520321846], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-BOTH", "prediction": 1, "probabilities": [0.07122167199850082, 0.9182708263397217, 0.010507442988455296], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-AR", "prediction": 1, "probabilities": [0.05922692269086838, 0.9350356459617615, 0.0057373931631445885], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-DIFFUSION", "prediction": 1, "probabilities": [0.09154275059700012, 0.8991622924804688, 0.009294936433434486], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-BOTH", "prediction": 1, "probabilities": [0.061033762991428375, 0.9314244389533997, 0.007541832514107227], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-AR", "prediction": 1, "probabilities": [0.15029682219028473, 0.8338066935539246, 0.015896489843726158], "tokens": 54} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-DIFFUSION", "prediction": 1, "probabilities": [0.1493491679430008, 0.8381385207176208, 0.012512396089732647], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-BOTH", "prediction": 1, "probabilities": [0.15806926786899567, 0.8166195750236511, 0.02531111054122448], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-AR", "prediction": 1, "probabilities": [0.12284184992313385, 0.8147609829902649, 0.06239711865782738], "tokens": 55} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-DIFFUSION", "prediction": 1, "probabilities": [0.15438716113567352, 0.8127452731132507, 0.03286757692694664], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-BOTH", "prediction": 1, "probabilities": [0.05095402151346207, 0.9151641726493835, 0.0338817834854126], "tokens": 51} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-AR", "prediction": 1, "probabilities": [0.012053163722157478, 0.9802168011665344, 0.007730037905275822], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-DIFFUSION", "prediction": 1, "probabilities": [0.00793106947094202, 0.9886552095413208, 0.0034137305337935686], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-BOTH", "prediction": 1, "probabilities": [0.015073616988956928, 0.9796390533447266, 0.0052872817032039165], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-AR", "prediction": 1, "probabilities": [0.0106365280225873, 0.9812145233154297, 0.008148976601660252], "tokens": 70} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-DIFFUSION", "prediction": 1, "probabilities": [0.012561365030705929, 0.9810971021652222, 0.006341456435620785], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-BOTH", "prediction": 1, "probabilities": [0.015676748007535934, 0.9767241477966309, 0.007599103730171919], "tokens": 69} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-AR", "prediction": 1, "probabilities": [0.026325054466724396, 0.9671263098716736, 0.006548679433763027], "tokens": 75} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-DIFFUSION", "prediction": 1, "probabilities": [0.013483992777764797, 0.9828107357025146, 0.003705312730744481], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-BOTH", "prediction": 1, "probabilities": [0.018457142636179924, 0.9755716323852539, 0.005971146747469902], "tokens": 72} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-AR", "prediction": 1, "probabilities": [0.019836224615573883, 0.9760860204696655, 0.004077753517776728], "tokens": 76} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-DIFFUSION", "prediction": 1, "probabilities": [0.011025538668036461, 0.9858829379081726, 0.0030914945527911186], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-BOTH", "prediction": 1, "probabilities": [0.030234958976507187, 0.9654324650764465, 0.004332525655627251], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-AR", "prediction": 1, "probabilities": [0.04499543458223343, 0.9481312036514282, 0.006873317528516054], "tokens": 75} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-DIFFUSION", "prediction": 1, "probabilities": [0.016597580164670944, 0.9790056347846985, 0.00439680740237236], "tokens": 69} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-BOTH", "prediction": 1, "probabilities": [0.03222670033574104, 0.9623064398765564, 0.005466865375638008], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-AR", "prediction": 1, "probabilities": [0.024785134941339493, 0.9649855494499207, 0.010229330509901047], "tokens": 76} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-DIFFUSION", "prediction": 1, "probabilities": [0.020922807976603508, 0.9692927002906799, 0.009784435853362083], "tokens": 72} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-BOTH", "prediction": 1, "probabilities": [0.01654072478413582, 0.9742736220359802, 0.009185664355754852], "tokens": 70} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-AR", "prediction": 1, "probabilities": [0.10888531059026718, 0.8623100519180298, 0.028804682195186615], "tokens": 34} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-DIFFUSION", "prediction": 1, "probabilities": [0.08413451164960861, 0.8822680115699768, 0.033597398549318314], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-BOTH", "prediction": 1, "probabilities": [0.198007270693779, 0.703702986240387, 0.09828975796699524], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-AR", "prediction": 1, "probabilities": [0.19076666235923767, 0.6480805277824402, 0.16115278005599976], "tokens": 33} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-DIFFUSION", "prediction": 1, "probabilities": [0.19651179015636444, 0.6741630434989929, 0.12932515144348145], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-BOTH", "prediction": 1, "probabilities": [0.13818621635437012, 0.784957766532898, 0.07685601711273193], "tokens": 28} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-AR", "prediction": 1, "probabilities": [0.25943782925605774, 0.7107425928115845, 0.029819536954164505], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-DIFFUSION", "prediction": 1, "probabilities": [0.1785583347082138, 0.7797240614891052, 0.04171758145093918], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-BOTH", "prediction": 1, "probabilities": [0.2044137716293335, 0.772135317325592, 0.023450858891010284], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-AR", "prediction": 1, "probabilities": [0.1304483860731125, 0.8633231520652771, 0.006228437647223473], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-DIFFUSION", "prediction": 1, "probabilities": [0.2852630615234375, 0.6936716437339783, 0.02106529287993908], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-BOTH", "prediction": 1, "probabilities": [0.21527869999408722, 0.7699573636054993, 0.014763905666768551], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-AR", "prediction": 1, "probabilities": [0.31733688712120056, 0.6514862179756165, 0.03117695450782776], "tokens": 33} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-DIFFUSION", "prediction": 1, "probabilities": [0.3948500454425812, 0.5843249559402466, 0.020825037732720375], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-BOTH", "prediction": 1, "probabilities": [0.42254066467285156, 0.525810718536377, 0.05164865031838417], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-AR", "prediction": 0, "probabilities": [0.4251426160335541, 0.4233537018299103, 0.15150368213653564], "tokens": 34} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-DIFFUSION", "prediction": 0, "probabilities": [0.5004169344902039, 0.42556560039520264, 0.0740174949169159], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-BOTH", "prediction": 1, "probabilities": [0.12756772339344025, 0.7038884162902832, 0.16854386031627655], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-AR", "prediction": 1, "probabilities": [0.060591306537389755, 0.922016441822052, 0.017392253503203392], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-DIFFUSION", "prediction": 1, "probabilities": [0.029716165736317635, 0.9614229202270508, 0.008860945701599121], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-BOTH", "prediction": 1, "probabilities": [0.061976220458745956, 0.9216965436935425, 0.016327250748872757], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-AR", "prediction": 1, "probabilities": [0.13958211243152618, 0.8189153075218201, 0.04150254651904106], "tokens": 32} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07408623397350311, 0.900846004486084, 0.025067675858736038], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-BOTH", "prediction": 1, "probabilities": [0.12014228105545044, 0.8145033121109009, 0.06535438448190689], "tokens": 31} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-AR", "prediction": 1, "probabilities": [0.2804009020328522, 0.7060931921005249, 0.013505898416042328], "tokens": 37} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-DIFFUSION", "prediction": 1, "probabilities": [0.07330591976642609, 0.9188506007194519, 0.007843482308089733], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-BOTH", "prediction": 1, "probabilities": [0.12453767657279968, 0.8464076519012451, 0.029054682701826096], "tokens": 34} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-AR", "prediction": 1, "probabilities": [0.19683676958084106, 0.7950892448425293, 0.00807395949959755], "tokens": 38} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-DIFFUSION", "prediction": 1, "probabilities": [0.11339495331048965, 0.871861457824707, 0.014743633568286896], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-BOTH", "prediction": 1, "probabilities": [0.24034684896469116, 0.7435397505760193, 0.016113409772515297], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-AR", "prediction": 0, "probabilities": [0.6043493151664734, 0.3659241795539856, 0.02972654066979885], "tokens": 37} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-DIFFUSION", "prediction": 1, "probabilities": [0.13724784553050995, 0.8450289964675903, 0.01772310957312584], "tokens": 31} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-BOTH", "prediction": 0, "probabilities": [0.6317940950393677, 0.3416476249694824, 0.026558255776762962], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-AR", "prediction": 1, "probabilities": [0.33419156074523926, 0.5673766732215881, 0.09843175113201141], "tokens": 38} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-DIFFUSION", "prediction": 1, "probabilities": [0.18995031714439392, 0.7471429109573364, 0.06290685385465622], "tokens": 34} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-BOTH", "prediction": 1, "probabilities": [0.15630792081356049, 0.7478838562965393, 0.095808245241642], "tokens": 32} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-AR", "prediction": 1, "probabilities": [0.037617504596710205, 0.9478644728660583, 0.014518044888973236], "tokens": 53} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-DIFFUSION", "prediction": 1, "probabilities": [0.017187193036079407, 0.9734005331993103, 0.009412233717739582], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-BOTH", "prediction": 1, "probabilities": [0.024699745699763298, 0.9620879888534546, 0.013212294317781925], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-AR", "prediction": 1, "probabilities": [0.026709139347076416, 0.9391898512840271, 0.034101054072380066], "tokens": 52} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-DIFFUSION", "prediction": 1, "probabilities": [0.02189885824918747, 0.9574487805366516, 0.020652323961257935], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-BOTH", "prediction": 1, "probabilities": [0.0106437336653471, 0.9755430817604065, 0.01381324790418148], "tokens": 47} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-AR", "prediction": 1, "probabilities": [0.03830554708838463, 0.951465368270874, 0.010229025967419147], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-DIFFUSION", "prediction": 1, "probabilities": [0.04226815328001976, 0.946029782295227, 0.011702079325914383], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-BOTH", "prediction": 1, "probabilities": [0.03367045521736145, 0.9601243138313293, 0.006205286365002394], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-AR", "prediction": 1, "probabilities": [0.010759089142084122, 0.9866878390312195, 0.0025530154816806316], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-DIFFUSION", "prediction": 1, "probabilities": [0.01948392577469349, 0.9755958318710327, 0.004920299164950848], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-BOTH", "prediction": 1, "probabilities": [0.018497519195079803, 0.9780498147010803, 0.0034526793751865625], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-AR", "prediction": 1, "probabilities": [0.056755438446998596, 0.9301331639289856, 0.013111338950693607], "tokens": 52} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-DIFFUSION", "prediction": 1, "probabilities": [0.048426587134599686, 0.9442560076713562, 0.007317410781979561], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-BOTH", "prediction": 1, "probabilities": [0.07879617065191269, 0.9068469405174255, 0.014356867410242558], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-AR", "prediction": 1, "probabilities": [0.023718418553471565, 0.9421862363815308, 0.034095339477062225], "tokens": 53} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-DIFFUSION", "prediction": 1, "probabilities": [0.028881371021270752, 0.955947756767273, 0.015170860104262829], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-BOTH", "prediction": 1, "probabilities": [0.00958208367228508, 0.9745240807533264, 0.015893761068582535], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-AR", "prediction": 1, "probabilities": [0.032387617975473404, 0.9596679210662842, 0.007944507524371147], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-DIFFUSION", "prediction": 1, "probabilities": [0.021595029160380363, 0.9744635224342346, 0.003941414877772331], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-BOTH", "prediction": 1, "probabilities": [0.06322866678237915, 0.9265391826629639, 0.010232084430754185], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-AR", "prediction": 1, "probabilities": [0.053042132407426834, 0.9258677363395691, 0.021090151742100716], "tokens": 43} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-DIFFUSION", "prediction": 1, "probabilities": [0.04339858889579773, 0.9460621476173401, 0.010539239272475243], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-BOTH", "prediction": 1, "probabilities": [0.06379850953817368, 0.9184580445289612, 0.017743460834026337], "tokens": 42} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-AR", "prediction": 1, "probabilities": [0.10538080334663391, 0.8880425691604614, 0.006576599553227425], "tokens": 48} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-DIFFUSION", "prediction": 1, "probabilities": [0.06288399547338486, 0.933384895324707, 0.0037311273626983166], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-BOTH", "prediction": 1, "probabilities": [0.0999516099691391, 0.889276385307312, 0.01077206153422594], "tokens": 45} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-AR", "prediction": 1, "probabilities": [0.07668948918581009, 0.91881263256073, 0.004497811198234558], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-DIFFUSION", "prediction": 1, "probabilities": [0.045792438089847565, 0.949802577495575, 0.004404998384416103], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-BOTH", "prediction": 1, "probabilities": [0.12950052320957184, 0.8645724654197693, 0.005926945712417364], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-AR", "prediction": 1, "probabilities": [0.20471762120723724, 0.7853856086730957, 0.009896761737763882], "tokens": 48} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-DIFFUSION", "prediction": 1, "probabilities": [0.04140826687216759, 0.9544214010238647, 0.0041703893803060055], "tokens": 42} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-BOTH", "prediction": 1, "probabilities": [0.36493292450904846, 0.6261367201805115, 0.008930394425988197], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-AR", "prediction": 1, "probabilities": [0.15295414626598358, 0.815432071685791, 0.031613789498806], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-DIFFUSION", "prediction": 1, "probabilities": [0.1166829913854599, 0.8573710322380066, 0.02594602480530739], "tokens": 45} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-BOTH", "prediction": 1, "probabilities": [0.08681900799274445, 0.8904359340667725, 0.022745048627257347], "tokens": 43} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-AR", "prediction": 1, "probabilities": [0.11727933585643768, 0.8237161040306091, 0.05900454893708229], "tokens": 30} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-DIFFUSION", "prediction": 1, "probabilities": [0.10420972108840942, 0.8252507448196411, 0.07053952664136887], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-BOTH", "prediction": 1, "probabilities": [0.15835531055927277, 0.6719227433204651, 0.16972194612026215], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-AR", "prediction": 1, "probabilities": [0.20682504773139954, 0.4642851650714874, 0.32888978719711304], "tokens": 29} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-DIFFUSION", "prediction": 1, "probabilities": [0.3193018138408661, 0.4085896611213684, 0.2721085250377655], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-BOTH", "prediction": 1, "probabilities": [0.2375795692205429, 0.560793936252594, 0.20162653923034668], "tokens": 24} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-AR", "prediction": 1, "probabilities": [0.1695641428232193, 0.7925500273704529, 0.03788583725690842], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-DIFFUSION", "prediction": 1, "probabilities": [0.17934085428714752, 0.7879043221473694, 0.0327548086643219], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-BOTH", "prediction": 1, "probabilities": [0.1724940538406372, 0.7899560332298279, 0.03754984587430954], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-AR", "prediction": 1, "probabilities": [0.07221546024084091, 0.9192095398902893, 0.008574939332902431], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1021994799375534, 0.8665438294410706, 0.03125665336847305], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-BOTH", "prediction": 1, "probabilities": [0.14110106229782104, 0.8061791658401489, 0.05271979793906212], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-AR", "prediction": 1, "probabilities": [0.4294095039367676, 0.5261509418487549, 0.044439565390348434], "tokens": 29} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-DIFFUSION", "prediction": 1, "probabilities": [0.4566153585910797, 0.5050846934318542, 0.03829994797706604], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-BOTH", "prediction": 0, "probabilities": [0.5127195119857788, 0.3980911374092102, 0.08918933570384979], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-AR", "prediction": 2, "probabilities": [0.29265204071998596, 0.34553414583206177, 0.36181387305259705], "tokens": 30} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-DIFFUSION", "prediction": 0, "probabilities": [0.44051244854927063, 0.4115926921367645, 0.14789490401744843], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-BOTH", "prediction": 1, "probabilities": [0.12573544681072235, 0.6197457909584045, 0.2545187771320343], "tokens": 26} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-AR", "prediction": 1, "probabilities": [0.05077727884054184, 0.9052577614784241, 0.043964989483356476], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-DIFFUSION", "prediction": 1, "probabilities": [0.012667556293308735, 0.9675268530845642, 0.019805585965514183], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-BOTH", "prediction": 1, "probabilities": [0.03125208616256714, 0.9284716844558716, 0.04027627408504486], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-AR", "prediction": 1, "probabilities": [0.1591413915157318, 0.7122370004653931, 0.12862154841423035], "tokens": 26} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.13663789629936218, 0.7542491555213928, 0.1091129258275032], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-BOTH", "prediction": 1, "probabilities": [0.13346803188323975, 0.7158846855163574, 0.15064726769924164], "tokens": 25} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-AR", "prediction": 1, "probabilities": [0.2994508743286133, 0.6825394630432129, 0.01800963468849659], "tokens": 31} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-DIFFUSION", "prediction": 1, "probabilities": [0.07711513340473175, 0.908222496509552, 0.014662377536296844], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-BOTH", "prediction": 1, "probabilities": [0.1671595573425293, 0.7665399312973022, 0.06630060076713562], "tokens": 28} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-AR", "prediction": 1, "probabilities": [0.14701321721076965, 0.8396798372268677, 0.013306980021297932], "tokens": 32} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1347080022096634, 0.8367667198181152, 0.028525277972221375], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-BOTH", "prediction": 1, "probabilities": [0.24558794498443604, 0.7219225168228149, 0.03248951584100723], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-AR", "prediction": 0, "probabilities": [0.6681478023529053, 0.2817584276199341, 0.05009371414780617], "tokens": 31} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-DIFFUSION", "prediction": 1, "probabilities": [0.09908226877450943, 0.869334876537323, 0.03158283978700638], "tokens": 25} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-BOTH", "prediction": 0, "probabilities": [0.7855879068374634, 0.17840717732906342, 0.036004919558763504], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-AR", "prediction": 1, "probabilities": [0.4155651032924652, 0.4172208905220032, 0.16721400618553162], "tokens": 32} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-DIFFUSION", "prediction": 1, "probabilities": [0.35914525389671326, 0.47479167580604553, 0.16606314480304718], "tokens": 28} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-BOTH", "prediction": 1, "probabilities": [0.24852187931537628, 0.4972797930240631, 0.2541983425617218], "tokens": 26} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-AR", "prediction": 1, "probabilities": [0.03415747731924057, 0.9506736397743225, 0.015168905258178711], "tokens": 58} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-DIFFUSION", "prediction": 1, "probabilities": [0.026376936584711075, 0.9619269371032715, 0.01169615425169468], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-BOTH", "prediction": 1, "probabilities": [0.017878344282507896, 0.9725104570388794, 0.009611125104129314], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-AR", "prediction": 1, "probabilities": [0.027323594316840172, 0.9433727264404297, 0.029303735122084618], "tokens": 57} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-DIFFUSION", "prediction": 1, "probabilities": [0.026944924145936966, 0.9488644599914551, 0.024190664291381836], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-BOTH", "prediction": 1, "probabilities": [0.020076178014278412, 0.9654501676559448, 0.01447368785738945], "tokens": 52} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-AR", "prediction": 1, "probabilities": [0.02930195815861225, 0.9604577422142029, 0.010240353643894196], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-DIFFUSION", "prediction": 1, "probabilities": [0.0326588898897171, 0.9537131190299988, 0.013628003187477589], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-BOTH", "prediction": 1, "probabilities": [0.028723418712615967, 0.9625885486602783, 0.008688006550073624], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-AR", "prediction": 1, "probabilities": [0.015448841266334057, 0.9802085161209106, 0.004342654719948769], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-DIFFUSION", "prediction": 1, "probabilities": [0.031857918947935104, 0.9602002501487732, 0.007941784337162971], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-BOTH", "prediction": 1, "probabilities": [0.02084924280643463, 0.9718044400215149, 0.007346317637711763], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-AR", "prediction": 1, "probabilities": [0.051097385585308075, 0.9348911643028259, 0.014011521823704243], "tokens": 57} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-DIFFUSION", "prediction": 1, "probabilities": [0.04097631573677063, 0.9491477608680725, 0.009875883348286152], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-BOTH", "prediction": 1, "probabilities": [0.04281587898731232, 0.9416301250457764, 0.015554011799395084], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-AR", "prediction": 1, "probabilities": [0.040171124041080475, 0.9294257164001465, 0.030403167009353638], "tokens": 58} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-DIFFUSION", "prediction": 1, "probabilities": [0.05092659965157509, 0.9296185374259949, 0.01945481449365616], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-BOTH", "prediction": 1, "probabilities": [0.01785781979560852, 0.9609829187393188, 0.02115929313004017], "tokens": 54} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-AR", "prediction": 1, "probabilities": [0.02099725231528282, 0.969376802444458, 0.009625932201743126], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-DIFFUSION", "prediction": 1, "probabilities": [0.01806853711605072, 0.9779125452041626, 0.004018992185592651], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-BOTH", "prediction": 1, "probabilities": [0.03221539407968521, 0.9604257345199585, 0.007358924485743046], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-AR", "prediction": 1, "probabilities": [0.032332196831703186, 0.9471373558044434, 0.020530441775918007], "tokens": 46} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-DIFFUSION", "prediction": 1, "probabilities": [0.040656741708517075, 0.947644829750061, 0.011698459275066853], "tokens": 44} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-BOTH", "prediction": 1, "probabilities": [0.040544506162405014, 0.9436522722244263, 0.015803193673491478], "tokens": 45} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-AR", "prediction": 1, "probabilities": [0.05138896405696869, 0.9406497478485107, 0.007961297407746315], "tokens": 50} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-DIFFUSION", "prediction": 1, "probabilities": [0.04074348881840706, 0.95467209815979, 0.0045844996348023415], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-BOTH", "prediction": 1, "probabilities": [0.05710400268435478, 0.934988260269165, 0.007907692342996597], "tokens": 47} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-AR", "prediction": 1, "probabilities": [0.04151242971420288, 0.9532833099365234, 0.005204233340919018], "tokens": 51} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-DIFFUSION", "prediction": 1, "probabilities": [0.02518724650144577, 0.970247209072113, 0.004565546754747629], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-BOTH", "prediction": 1, "probabilities": [0.08234621584415436, 0.9114890694618225, 0.006164702586829662], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-AR", "prediction": 1, "probabilities": [0.15509115159511566, 0.8330848813056946, 0.011823956854641438], "tokens": 50} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-DIFFUSION", "prediction": 1, "probabilities": [0.04194055497646332, 0.9522800445556641, 0.005779379978775978], "tokens": 44} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-BOTH", "prediction": 1, "probabilities": [0.17613494396209717, 0.8149424195289612, 0.008922658860683441], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-AR", "prediction": 1, "probabilities": [0.10237694531679153, 0.8638429045677185, 0.033780165016651154], "tokens": 52} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-DIFFUSION", "prediction": 1, "probabilities": [0.09634120017290115, 0.8798030614852905, 0.02385574020445347], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-BOTH", "prediction": 1, "probabilities": [0.0711035504937172, 0.9036721587181091, 0.025224294513463974], "tokens": 46} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-AR", "prediction": 1, "probabilities": [0.04538707807660103, 0.9352543950080872, 0.019358554854989052], "tokens": 59} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION", "prediction": 1, "probabilities": [0.02421022579073906, 0.963483452796936, 0.01230631023645401], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-BOTH", "prediction": 1, "probabilities": [0.023030756041407585, 0.9648289680480957, 0.012140302918851376], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-AR", "prediction": 1, "probabilities": [0.035967208445072174, 0.9173038601875305, 0.04672889783978462], "tokens": 58} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-DIFFUSION", "prediction": 1, "probabilities": [0.02281099744141102, 0.9541978240013123, 0.022991133853793144], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-BOTH", "prediction": 1, "probabilities": [0.019920509308576584, 0.9658147096633911, 0.014264780096709728], "tokens": 53} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-AR", "prediction": 1, "probabilities": [0.049783460795879364, 0.9368708729743958, 0.013345685787498951], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-DIFFUSION", "prediction": 1, "probabilities": [0.031346678733825684, 0.9568292498588562, 0.011824066750705242], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-BOTH", "prediction": 1, "probabilities": [0.030521102249622345, 0.9618420004844666, 0.0076368716545403], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-AR", "prediction": 1, "probabilities": [0.02000441774725914, 0.9743998050689697, 0.0055957576259970665], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-DIFFUSION", "prediction": 1, "probabilities": [0.025151193141937256, 0.9667106866836548, 0.008138121105730534], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-BOTH", "prediction": 1, "probabilities": [0.018966946750879288, 0.9749979972839355, 0.00603507598862052], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-AR", "prediction": 1, "probabilities": [0.05566444247961044, 0.9309051632881165, 0.013430414721369743], "tokens": 58} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-DIFFUSION", "prediction": 1, "probabilities": [0.03949468210339546, 0.9521523714065552, 0.008352934382855892], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-BOTH", "prediction": 1, "probabilities": [0.05976478010416031, 0.926087498664856, 0.014147769659757614], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-AR", "prediction": 1, "probabilities": [0.031106362119317055, 0.9372007846832275, 0.03169278800487518], "tokens": 59} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-DIFFUSION", "prediction": 1, "probabilities": [0.0494801364839077, 0.9285897612571716, 0.02193005196750164], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-BOTH", "prediction": 1, "probabilities": [0.01720908284187317, 0.9578604102134705, 0.024930564686655998], "tokens": 55} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-AR", "prediction": 1, "probabilities": [0.08416927605867386, 0.8757123351097107, 0.04011842608451843], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-DIFFUSION", "prediction": 1, "probabilities": [0.04507146775722504, 0.9356935620307922, 0.01923496276140213], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-BOTH", "prediction": 1, "probabilities": [0.08319851011037827, 0.870269775390625, 0.04653168097138405], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-AR", "prediction": 1, "probabilities": [0.2117764949798584, 0.687911331653595, 0.10031217336654663], "tokens": 35} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-DIFFUSION", "prediction": 1, "probabilities": [0.14745613932609558, 0.7799317240715027, 0.07261212170124054], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-BOTH", "prediction": 1, "probabilities": [0.22511421144008636, 0.661359965801239, 0.11352580785751343], "tokens": 34} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-AR", "prediction": 1, "probabilities": [0.37058359384536743, 0.6028667688369751, 0.02654963731765747], "tokens": 40} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-DIFFUSION", "prediction": 1, "probabilities": [0.1865207701921463, 0.7879366874694824, 0.025542540475726128], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-BOTH", "prediction": 1, "probabilities": [0.19018922746181488, 0.7278261780738831, 0.08198461681604385], "tokens": 37} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-AR", "prediction": 1, "probabilities": [0.2349810153245926, 0.7446154952049255, 0.020403524860739708], "tokens": 41} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-DIFFUSION", "prediction": 1, "probabilities": [0.26201826333999634, 0.7003011703491211, 0.037680577486753464], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-BOTH", "prediction": 1, "probabilities": [0.36675211787223816, 0.5972028970718384, 0.03604495897889137], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-AR", "prediction": 0, "probabilities": [0.676588773727417, 0.285077840089798, 0.03833335265517235], "tokens": 40} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-DIFFUSION", "prediction": 1, "probabilities": [0.22455871105194092, 0.7372704744338989, 0.03817083314061165], "tokens": 34} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-BOTH", "prediction": 0, "probabilities": [0.7345373630523682, 0.22438372671604156, 0.04107890650629997], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-AR", "prediction": 1, "probabilities": [0.4235791265964508, 0.43523311614990234, 0.14118771255016327], "tokens": 41} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-DIFFUSION", "prediction": 1, "probabilities": [0.4034234285354614, 0.47054699063301086, 0.12602952122688293], "tokens": 37} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-BOTH", "prediction": 1, "probabilities": [0.30024147033691406, 0.5226351618766785, 0.17712341248989105], "tokens": 35} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-AR", "prediction": 1, "probabilities": [0.0943385437130928, 0.8871569037437439, 0.018504543229937553], "tokens": 31} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-DIFFUSION", "prediction": 1, "probabilities": [0.09261453151702881, 0.8933359980583191, 0.014049443416297436], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-BOTH", "prediction": 1, "probabilities": [0.10893452912569046, 0.8500193357467651, 0.04104608669877052], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-AR", "prediction": 1, "probabilities": [0.17349539697170258, 0.6820238828659058, 0.14448069036006927], "tokens": 30} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-DIFFUSION", "prediction": 1, "probabilities": [0.32018837332725525, 0.5597260594367981, 0.12008553743362427], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-BOTH", "prediction": 1, "probabilities": [0.09289709478616714, 0.8631539940834045, 0.04394889995455742], "tokens": 25} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-AR", "prediction": 1, "probabilities": [0.22709906101226807, 0.7593816518783569, 0.013519352301955223], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-DIFFUSION", "prediction": 1, "probabilities": [0.14605236053466797, 0.8423076272010803, 0.011640067212283611], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-BOTH", "prediction": 1, "probabilities": [0.16301164031028748, 0.8288237452507019, 0.008164667524397373], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-AR", "prediction": 1, "probabilities": [0.0650075152516365, 0.9330529570579529, 0.001939514302648604], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-DIFFUSION", "prediction": 1, "probabilities": [0.12490148842334747, 0.8682776093482971, 0.006820943206548691], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-BOTH", "prediction": 1, "probabilities": [0.08924905955791473, 0.9025899767875671, 0.008160988800227642], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-AR", "prediction": 1, "probabilities": [0.33712559938430786, 0.6462928056716919, 0.016581634059548378], "tokens": 30} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-DIFFUSION", "prediction": 1, "probabilities": [0.3463335633277893, 0.646295428276062, 0.007370982319116592], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-BOTH", "prediction": 1, "probabilities": [0.43909308314323425, 0.5323833227157593, 0.02852359414100647], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-AR", "prediction": 1, "probabilities": [0.34422847628593445, 0.5378547310829163, 0.1179167851805687], "tokens": 31} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-DIFFUSION", "prediction": 1, "probabilities": [0.45327216386795044, 0.5040689706802368, 0.04265886917710304], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-BOTH", "prediction": 1, "probabilities": [0.0652901753783226, 0.8693559169769287, 0.06535390019416809], "tokens": 27} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-AR", "prediction": 1, "probabilities": [0.06082341447472572, 0.9300304651260376, 0.009146097116172314], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-DIFFUSION", "prediction": 1, "probabilities": [0.040339190512895584, 0.9553325772285461, 0.004328235052525997], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-BOTH", "prediction": 1, "probabilities": [0.0984986424446106, 0.8939651846885681, 0.007536136545240879], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-AR", "prediction": 1, "probabilities": [0.11450106650590897, 0.8644624352455139, 0.021036438643932343], "tokens": 42} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-DIFFUSION", "prediction": 1, "probabilities": [0.12699724733829498, 0.8622111082077026, 0.01079165656119585], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-BOTH", "prediction": 1, "probabilities": [0.14580148458480835, 0.8337706923484802, 0.020427782088518143], "tokens": 41} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-AR", "prediction": 1, "probabilities": [0.21295662224292755, 0.7784897089004517, 0.008553678169846535], "tokens": 47} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-DIFFUSION", "prediction": 1, "probabilities": [0.20392447710037231, 0.7908311486244202, 0.005244371015578508], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-BOTH", "prediction": 1, "probabilities": [0.22761543095111847, 0.7606860399246216, 0.011698535643517971], "tokens": 44} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-AR", "prediction": 1, "probabilities": [0.15668430924415588, 0.838441014289856, 0.00487470394000411], "tokens": 48} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-DIFFUSION", "prediction": 1, "probabilities": [0.1626823991537094, 0.8321754932403564, 0.00514207361266017], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-BOTH", "prediction": 1, "probabilities": [0.2594612240791321, 0.735164225101471, 0.005374548025429249], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-AR", "prediction": 1, "probabilities": [0.48892122507095337, 0.5014709234237671, 0.00960785523056984], "tokens": 47} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-DIFFUSION", "prediction": 1, "probabilities": [0.15693345665931702, 0.8366048336029053, 0.006461726501584053], "tokens": 41} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-BOTH", "prediction": 0, "probabilities": [0.6747605204582214, 0.32006800174713135, 0.00517140980809927], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-AR", "prediction": 1, "probabilities": [0.3020084500312805, 0.6705597639083862, 0.027431830763816833], "tokens": 48} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-DIFFUSION", "prediction": 1, "probabilities": [0.2893812954425812, 0.6917532682418823, 0.018865425139665604], "tokens": 44} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-BOTH", "prediction": 1, "probabilities": [0.18708014488220215, 0.7916812896728516, 0.021238604560494423], "tokens": 42} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-AR", "prediction": 1, "probabilities": [0.08660994470119476, 0.90426105260849, 0.0091289933770895], "tokens": 62} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-DIFFUSION", "prediction": 1, "probabilities": [0.06718919426202774, 0.9251578450202942, 0.007652913220226765], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-BOTH", "prediction": 1, "probabilities": [0.06342107802629471, 0.9291912913322449, 0.007387596648186445], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-AR", "prediction": 1, "probabilities": [0.06095373257994652, 0.9155647158622742, 0.023481590673327446], "tokens": 61} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07926391810178757, 0.905142068862915, 0.015593983232975006], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-BOTH", "prediction": 1, "probabilities": [0.06606888025999069, 0.9189593195915222, 0.014971818774938583], "tokens": 56} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-AR", "prediction": 1, "probabilities": [0.06939492374658585, 0.9231874942779541, 0.007417517248541117], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-DIFFUSION", "prediction": 1, "probabilities": [0.07203836739063263, 0.9191924333572388, 0.008769204840064049], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-BOTH", "prediction": 1, "probabilities": [0.055904898792505264, 0.9389883279800415, 0.005106777884066105], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-AR", "prediction": 1, "probabilities": [0.032078951597213745, 0.9654738306999207, 0.0024472465738654137], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-DIFFUSION", "prediction": 1, "probabilities": [0.06849054247140884, 0.9267265200614929, 0.004782878793776035], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-BOTH", "prediction": 1, "probabilities": [0.044044286012649536, 0.9527239799499512, 0.0032317182049155235], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-AR", "prediction": 1, "probabilities": [0.10788837820291519, 0.8835031390190125, 0.008608510717749596], "tokens": 61} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-DIFFUSION", "prediction": 1, "probabilities": [0.11513914167881012, 0.8796746134757996, 0.005186211317777634], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-BOTH", "prediction": 1, "probabilities": [0.11084125190973282, 0.8814991116523743, 0.007659710478037596], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-AR", "prediction": 1, "probabilities": [0.07856975495815277, 0.8992480039596558, 0.022182265296578407], "tokens": 62} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-DIFFUSION", "prediction": 1, "probabilities": [0.109621062874794, 0.8779095411300659, 0.012469478882849216], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-BOTH", "prediction": 1, "probabilities": [0.03706672415137291, 0.9495589733123779, 0.013374360278248787], "tokens": 58} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-AR", "prediction": 1, "probabilities": [0.03699536994099617, 0.9549511671066284, 0.008053447119891644], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-DIFFUSION", "prediction": 1, "probabilities": [0.021840421482920647, 0.9748489260673523, 0.003310642670840025], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-BOTH", "prediction": 1, "probabilities": [0.04247281327843666, 0.9517195224761963, 0.005807641427963972], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR", "prediction": 1, "probabilities": [0.05638135224580765, 0.9249981045722961, 0.018620500341057777], "tokens": 43} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-DIFFUSION", "prediction": 1, "probabilities": [0.04191704839468002, 0.9471937417984009, 0.010889283381402493], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-BOTH", "prediction": 1, "probabilities": [0.06882010400295258, 0.9136105179786682, 0.01756935566663742], "tokens": 42} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-AR", "prediction": 1, "probabilities": [0.08673319965600967, 0.9088189601898193, 0.004447883926331997], "tokens": 48} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-DIFFUSION", "prediction": 1, "probabilities": [0.04384264349937439, 0.9526275992393494, 0.0035297677386552095], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-BOTH", "prediction": 1, "probabilities": [0.07571221888065338, 0.9151041507720947, 0.009183642454445362], "tokens": 45} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-AR", "prediction": 1, "probabilities": [0.08401229977607727, 0.9116314053535461, 0.004356237594038248], "tokens": 49} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-DIFFUSION", "prediction": 1, "probabilities": [0.04122069850564003, 0.9540950655937195, 0.004684277344495058], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-BOTH", "prediction": 1, "probabilities": [0.15328845381736755, 0.8419452905654907, 0.0047662402503192425], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-AR", "prediction": 1, "probabilities": [0.28529974818229675, 0.7062467336654663, 0.008453484624624252], "tokens": 48} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-DIFFUSION", "prediction": 1, "probabilities": [0.05200691521167755, 0.9419975876808167, 0.005995551589876413], "tokens": 42} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-BOTH", "prediction": 1, "probabilities": [0.35241103172302246, 0.6385523080825806, 0.00903666764497757], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-AR", "prediction": 1, "probabilities": [0.1500212699174881, 0.8220308423042297, 0.027947859838604927], "tokens": 49} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-DIFFUSION", "prediction": 1, "probabilities": [0.09403843432664871, 0.8843205571174622, 0.021640989929437637], "tokens": 45} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-BOTH", "prediction": 1, "probabilities": [0.08601763844490051, 0.8879279494285583, 0.02605440653860569], "tokens": 43} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-AR", "prediction": 1, "probabilities": [0.11190266907215118, 0.8707151412963867, 0.017382191494107246], "tokens": 41} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-DIFFUSION", "prediction": 1, "probabilities": [0.12782633304595947, 0.8591827750205994, 0.012990852817893028], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-BOTH", "prediction": 1, "probabilities": [0.10569276660680771, 0.87203449010849, 0.022272733971476555], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-AR", "prediction": 1, "probabilities": [0.15947137773036957, 0.7356976270675659, 0.10483105480670929], "tokens": 40} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-DIFFUSION", "prediction": 1, "probabilities": [0.2466246783733368, 0.6943180561065674, 0.05905727297067642], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-BOTH", "prediction": 1, "probabilities": [0.12558230757713318, 0.8415787816047668, 0.032838959246873856], "tokens": 35} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-AR", "prediction": 1, "probabilities": [0.1433747559785843, 0.8446434140205383, 0.011981801129877567], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-DIFFUSION", "prediction": 1, "probabilities": [0.15361705422401428, 0.833138644695282, 0.013244322501122952], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-BOTH", "prediction": 1, "probabilities": [0.13784493505954742, 0.849295437335968, 0.012859627604484558], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-AR", "prediction": 1, "probabilities": [0.06920261681079865, 0.9278473854064941, 0.002950004767626524], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-DIFFUSION", "prediction": 1, "probabilities": [0.10623232275247574, 0.8844868540763855, 0.009280814789235592], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-BOTH", "prediction": 1, "probabilities": [0.11323604732751846, 0.8751441240310669, 0.011619853787124157], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-AR", "prediction": 1, "probabilities": [0.2584046423435211, 0.7252143621444702, 0.016381027176976204], "tokens": 40} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-DIFFUSION", "prediction": 1, "probabilities": [0.23677153885364532, 0.7535242438316345, 0.009704172611236572], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-BOTH", "prediction": 1, "probabilities": [0.2512330412864685, 0.7253256440162659, 0.023441338911652565], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-AR", "prediction": 1, "probabilities": [0.2713022530078888, 0.6459911465644836, 0.08270659297704697], "tokens": 41} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-DIFFUSION", "prediction": 1, "probabilities": [0.2860986292362213, 0.6820154190063477, 0.031885989010334015], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-BOTH", "prediction": 1, "probabilities": [0.10158395767211914, 0.8370151519775391, 0.06140084192156792], "tokens": 37} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-AR", "prediction": 1, "probabilities": [0.037855129688978195, 0.9533042907714844, 0.008840526454150677], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-DIFFUSION", "prediction": 1, "probabilities": [0.026794588193297386, 0.9687860608100891, 0.004419359378516674], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-BOTH", "prediction": 1, "probabilities": [0.06778807938098907, 0.9221128225326538, 0.010099144652485847], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-AR", "prediction": 1, "probabilities": [0.07824281603097916, 0.9017121195793152, 0.020045042037963867], "tokens": 37} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-DIFFUSION", "prediction": 1, "probabilities": [0.06507186591625214, 0.9241763949394226, 0.010751692578196526], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-BOTH", "prediction": 1, "probabilities": [0.07754477858543396, 0.9080321192741394, 0.014423124492168427], "tokens": 36} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-AR", "prediction": 1, "probabilities": [0.11503126472234726, 0.8782778382301331, 0.006690884940326214], "tokens": 42} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-DIFFUSION", "prediction": 1, "probabilities": [0.06720709800720215, 0.9294806122779846, 0.0033122722525149584], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-BOTH", "prediction": 1, "probabilities": [0.10598710179328918, 0.8829420208930969, 0.011070835404098034], "tokens": 39} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-AR", "prediction": 1, "probabilities": [0.08498828858137131, 0.9105163216590881, 0.00449536694213748], "tokens": 43} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-DIFFUSION", "prediction": 1, "probabilities": [0.06194743514060974, 0.933841347694397, 0.004211254417896271], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-BOTH", "prediction": 1, "probabilities": [0.16680987179279327, 0.8265311121940613, 0.0066590323112905025], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-AR", "prediction": 1, "probabilities": [0.27360719442367554, 0.7155676484107971, 0.01082519069314003], "tokens": 42} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-DIFFUSION", "prediction": 1, "probabilities": [0.04980207234621048, 0.9451349973678589, 0.005062876269221306], "tokens": 36} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-BOTH", "prediction": 1, "probabilities": [0.36003392934799194, 0.6295066475868225, 0.010459378361701965], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-AR", "prediction": 1, "probabilities": [0.18099220097064972, 0.7956801652908325, 0.02332761138677597], "tokens": 43} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-DIFFUSION", "prediction": 1, "probabilities": [0.15895839035511017, 0.8198549747467041, 0.02118656411767006], "tokens": 39} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-BOTH", "prediction": 1, "probabilities": [0.11519947648048401, 0.8655668497085571, 0.019233698025345802], "tokens": 37} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-AR", "prediction": 1, "probabilities": [0.06006286293268204, 0.9248092174530029, 0.015127953141927719], "tokens": 53} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-DIFFUSION", "prediction": 1, "probabilities": [0.04519784450531006, 0.9475085735321045, 0.007293567061424255], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-BOTH", "prediction": 1, "probabilities": [0.04042457789182663, 0.9495999217033386, 0.009975552558898926], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-AR", "prediction": 1, "probabilities": [0.07326894998550415, 0.8680711388587952, 0.05865989252924919], "tokens": 52} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0499405674636364, 0.9258393049240112, 0.024220142513513565], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-BOTH", "prediction": 1, "probabilities": [0.031079545617103577, 0.950798511505127, 0.018121974542737007], "tokens": 47} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-AR", "prediction": 1, "probabilities": [0.05600322410464287, 0.9340574145317078, 0.009939329698681831], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-DIFFUSION", "prediction": 1, "probabilities": [0.0473298653960228, 0.945556104183197, 0.00711404113098979], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-BOTH", "prediction": 1, "probabilities": [0.0500217080116272, 0.9433233737945557, 0.006654931232333183], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-AR", "prediction": 1, "probabilities": [0.030217990279197693, 0.9674996733665466, 0.002282405272126198], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-DIFFUSION", "prediction": 1, "probabilities": [0.04910306632518768, 0.9465779662132263, 0.004318928811699152], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-BOTH", "prediction": 1, "probabilities": [0.031755391508340836, 0.9639033675193787, 0.004341293592005968], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-AR", "prediction": 1, "probabilities": [0.08546324819326401, 0.903526246547699, 0.011010462418198586], "tokens": 52} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-DIFFUSION", "prediction": 1, "probabilities": [0.08125534653663635, 0.9130651354789734, 0.00567949702963233], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-BOTH", "prediction": 1, "probabilities": [0.0885152742266655, 0.897148072719574, 0.014336691237986088], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-AR", "prediction": 1, "probabilities": [0.07138838618993759, 0.8868001699447632, 0.04181142523884773], "tokens": 53} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-DIFFUSION", "prediction": 1, "probabilities": [0.0787501111626625, 0.8959106802940369, 0.0253391582518816], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-BOTH", "prediction": 1, "probabilities": [0.021612856537103653, 0.9457179307937622, 0.03266922011971474], "tokens": 49} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-AR", "prediction": 1, "probabilities": [0.010635746642947197, 0.9844058752059937, 0.004958337172865868], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-DIFFUSION", "prediction": 1, "probabilities": [0.012080841697752476, 0.985288679599762, 0.002630473580211401], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-BOTH", "prediction": 1, "probabilities": [0.02463037520647049, 0.9702326059341431, 0.0051370179280638695], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-AR", "prediction": 1, "probabilities": [0.016669539734721184, 0.9735904335975647, 0.00973997637629509], "tokens": 64} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0154871242120862, 0.9790054559707642, 0.005507404450327158], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-BOTH", "prediction": 1, "probabilities": [0.02263750322163105, 0.9673179984092712, 0.010044500231742859], "tokens": 63} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-AR", "prediction": 1, "probabilities": [0.031229183077812195, 0.9634345173835754, 0.00533634377643466], "tokens": 69} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-DIFFUSION", "prediction": 1, "probabilities": [0.022235147655010223, 0.9747800230979919, 0.0029847684781998396], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-BOTH", "prediction": 1, "probabilities": [0.026717720553278923, 0.9685863852500916, 0.004695884417742491], "tokens": 66} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-AR", "prediction": 1, "probabilities": [0.025270916521549225, 0.9712499380111694, 0.0034791298676282167], "tokens": 70} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-DIFFUSION", "prediction": 1, "probabilities": [0.01691845990717411, 0.9798098802566528, 0.0032716814894229174], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-BOTH", "prediction": 1, "probabilities": [0.03734446316957474, 0.9592143893241882, 0.0034411917440593243], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-AR", "prediction": 1, "probabilities": [0.07015746086835861, 0.9231408834457397, 0.006701616104692221], "tokens": 69} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-DIFFUSION", "prediction": 1, "probabilities": [0.023942524567246437, 0.9721002578735352, 0.003957155626267195], "tokens": 63} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-BOTH", "prediction": 1, "probabilities": [0.06250566244125366, 0.932346761226654, 0.005147567018866539], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-AR", "prediction": 1, "probabilities": [0.03724813088774681, 0.9478840827941895, 0.014867854304611683], "tokens": 70} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-DIFFUSION", "prediction": 1, "probabilities": [0.03405233472585678, 0.9554759860038757, 0.010471736080944538], "tokens": 66} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-BOTH", "prediction": 1, "probabilities": [0.025549141690135002, 0.9639240503311157, 0.010526787489652634], "tokens": 64} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-AR", "prediction": 1, "probabilities": [0.06409315764904022, 0.9094409942626953, 0.0264658872038126], "tokens": 46} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-DIFFUSION", "prediction": 1, "probabilities": [0.030834896489977837, 0.9523953795433044, 0.016769705340266228], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-BOTH", "prediction": 1, "probabilities": [0.030571676790714264, 0.9451302289962769, 0.024298114702105522], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-AR", "prediction": 1, "probabilities": [0.051121924072504044, 0.8642809391021729, 0.0845971331000328], "tokens": 45} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-DIFFUSION", "prediction": 1, "probabilities": [0.07382363826036453, 0.850994348526001, 0.07518196105957031], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-BOTH", "prediction": 1, "probabilities": [0.033740077167749405, 0.9190420508384705, 0.047217823565006256], "tokens": 40} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-AR", "prediction": 1, "probabilities": [0.05792364478111267, 0.9226797819137573, 0.019396597519516945], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-DIFFUSION", "prediction": 1, "probabilities": [0.05306778475642204, 0.9232560992240906, 0.023676153272390366], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-BOTH", "prediction": 1, "probabilities": [0.051928482949733734, 0.933384358882904, 0.014687206596136093], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-AR", "prediction": 1, "probabilities": [0.024451930075883865, 0.969613254070282, 0.005934841465204954], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-DIFFUSION", "prediction": 1, "probabilities": [0.042887527495622635, 0.9432199597358704, 0.013892513699829578], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-BOTH", "prediction": 1, "probabilities": [0.025547794997692108, 0.9660625457763672, 0.008389627560973167], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-AR", "prediction": 1, "probabilities": [0.08399491757154465, 0.8916747570037842, 0.02433028444647789], "tokens": 45} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-DIFFUSION", "prediction": 1, "probabilities": [0.11341767758131027, 0.868455708026886, 0.0181265901774168], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-BOTH", "prediction": 1, "probabilities": [0.10775203257799149, 0.8658536076545715, 0.02639436535537243], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-AR", "prediction": 1, "probabilities": [0.059486061334609985, 0.8606964945793152, 0.07981743663549423], "tokens": 46} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-DIFFUSION", "prediction": 1, "probabilities": [0.09620607644319534, 0.847745418548584, 0.05604846030473709], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-BOTH", "prediction": 1, "probabilities": [0.027191435918211937, 0.9122016429901123, 0.060606975108385086], "tokens": 42} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-AR", "prediction": 1, "probabilities": [0.028681818395853043, 0.9618858098983765, 0.009432393126189709], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-DIFFUSION", "prediction": 1, "probabilities": [0.01369484793394804, 0.9812065958976746, 0.005098602268844843], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-BOTH", "prediction": 1, "probabilities": [0.022606339305639267, 0.9688439965248108, 0.008549639023840427], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-AR", "prediction": 1, "probabilities": [0.057179756462574005, 0.9192516803741455, 0.023568622767925262], "tokens": 46} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-DIFFUSION", "prediction": 1, "probabilities": [0.03861670941114426, 0.9451157450675964, 0.01626756228506565], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-BOTH", "prediction": 1, "probabilities": [0.040128227323293686, 0.9371336102485657, 0.0227381382137537], "tokens": 45} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-AR", "prediction": 1, "probabilities": [0.07316895574331284, 0.9203481674194336, 0.00648284750059247], "tokens": 51} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-DIFFUSION", "prediction": 1, "probabilities": [0.04924684762954712, 0.9444904327392578, 0.0062627363950014114], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-BOTH", "prediction": 1, "probabilities": [0.05006186664104462, 0.9377143383026123, 0.012223721481859684], "tokens": 48} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-AR", "prediction": 1, "probabilities": [0.058296557515859604, 0.9371827244758606, 0.004520730581134558], "tokens": 52} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-DIFFUSION", "prediction": 1, "probabilities": [0.047626789659261703, 0.9456255435943604, 0.00674772821366787], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-BOTH", "prediction": 1, "probabilities": [0.07486715167760849, 0.9173138737678528, 0.007818943820893764], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-AR", "prediction": 1, "probabilities": [0.18345119059085846, 0.8029072284698486, 0.013641562312841415], "tokens": 51} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-DIFFUSION", "prediction": 1, "probabilities": [0.0514029785990715, 0.9403693675994873, 0.008227621205151081], "tokens": 45} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-BOTH", "prediction": 1, "probabilities": [0.2057662010192871, 0.7788276672363281, 0.015406136400997639], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-AR", "prediction": 1, "probabilities": [0.11486408859491348, 0.8508700728416443, 0.034265805035829544], "tokens": 52} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-DIFFUSION", "prediction": 1, "probabilities": [0.07032066583633423, 0.9079911708831787, 0.021688180044293404], "tokens": 48} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-BOTH", "prediction": 1, "probabilities": [0.07212971895933151, 0.8938366770744324, 0.03403358906507492], "tokens": 46} +{"id": "modality-text-global-zh-philosophy/test/165", "prediction": 0, "probabilities": [0.9988527297973633, 4.6894169827282894e-06, 0.0011426761047914624], "tokens": 65} +{"id": "modality-text-global-es-human_aging/test/127", "prediction": 0, "probabilities": [0.9992419481277466, 0.00018084167095366865, 0.0005772360018454492], "tokens": 90} +{"id": "modality-text-global-zh-elementary_mathematics/test/289", "prediction": 0, "probabilities": [0.9980288147926331, 0.0002217188011854887, 0.0017494397470727563], "tokens": 75} +{"id": "modality-text-global-ja-professional_law/test/570", "prediction": 2, "probabilities": [0.30941253900527954, 0.001302422839216888, 0.6892849802970886], "tokens": 298} +{"id": "modality-text-global-fr-professional_psychology/test/359", "prediction": 0, "probabilities": [0.9939032196998596, 0.004265942145138979, 0.0018307596910744905], "tokens": 51} +{"id": "modality-text-global-fr-logical_fallacies/test/136", "prediction": 0, "probabilities": [0.9953902959823608, 0.001135542057454586, 0.0034741421695798635], "tokens": 149} +{"id": "modality-text-global-de-moral_disputes/test/116", "prediction": 0, "probabilities": [0.9988677501678467, 0.0003212698793504387, 0.0008110075141303241], "tokens": 95} +{"id": "modality-text-global-ja-high_school_physics/test/67", "prediction": 0, "probabilities": [0.9997852444648743, 3.8727722539988463e-07, 0.00021435273811221123], "tokens": 68} +{"id": "modality-text-global-es-high_school_microeconomics/test/172", "prediction": 0, "probabilities": [0.9999357461929321, 3.201544677722268e-05, 3.2146846933756024e-05], "tokens": 52} +{"id": "modality-text-global-fr-high_school_microeconomics/test/123", "prediction": 0, "probabilities": [0.9998931884765625, 7.939196075312793e-05, 2.7432626666268334e-05], "tokens": 79} +{"id": "modality-text-global-es-professional_law/test/986", "prediction": 0, "probabilities": [0.9980379939079285, 0.0013203765265643597, 0.0006416256073862314], "tokens": 313} +{"id": "modality-text-global-es-prehistory/test/47", "prediction": 0, "probabilities": [0.9866119623184204, 0.011047884821891785, 0.0023401211947202682], "tokens": 100} +{"id": "modality-text-global-zh-professional_medicine/test/56", "prediction": 0, "probabilities": [0.647834062576294, 0.004639737773686647, 0.3475261628627777], "tokens": 126} +{"id": "modality-text-global-es-miscellaneous/test/639", "prediction": 0, "probabilities": [0.9989288449287415, 0.0008277379674836993, 0.00024337029026355594], "tokens": 47} +{"id": "modality-text-global-es-business_ethics/test/42", "prediction": 0, "probabilities": [0.9995895028114319, 0.000229867291636765, 0.0001806383952498436], "tokens": 60} +{"id": "modality-text-global-en-virology/test/107", "prediction": 0, "probabilities": [0.4441407024860382, 0.1770082712173462, 0.3788509666919708], "tokens": 125} +{"id": "modality-text-global-en-moral_scenarios/test/304", "prediction": 0, "probabilities": [0.9932032823562622, 0.00580384349450469, 0.0009928419021889567], "tokens": 98} +{"id": "modality-text-global-es-high_school_microeconomics/test/157", "prediction": 0, "probabilities": [0.999913215637207, 7.813621050445363e-05, 8.622827408544254e-06], "tokens": 76} +{"id": "modality-text-global-ja-moral_scenarios/test/453", "prediction": 0, "probabilities": [0.9853808283805847, 0.0006843189476057887, 0.013934880495071411], "tokens": 123} +{"id": "modality-text-global-de-high_school_statistics/test/185", "prediction": 0, "probabilities": [0.9999892711639404, 4.765070116263814e-06, 5.959917416475946e-06], "tokens": 208} +{"id": "modality-text-global-ja-moral_scenarios/test/73", "prediction": 0, "probabilities": [0.9931322932243347, 0.0006990740075707436, 0.006168621592223644], "tokens": 123} +{"id": "modality-text-global-fr-high_school_macroeconomics/test/270", "prediction": 0, "probabilities": [0.9980778694152832, 0.0018570938846096396, 6.505290366476402e-05], "tokens": 126} +{"id": "modality-text-global-en-professional_law/test/539", "prediction": 0, "probabilities": [0.7265213131904602, 0.039083946496248245, 0.23439471423625946], "tokens": 295} +{"id": "modality-text-global-de-high_school_macroeconomics/test/172", "prediction": 0, "probabilities": [0.9999300241470337, 2.3672288079978898e-05, 4.6298213419504464e-05], "tokens": 54} +{"id": "modality-text-global-en-professional_law/test/9", "prediction": 0, "probabilities": [0.6646255850791931, 0.17185312509536743, 0.16352126002311707], "tokens": 274} +{"id": "modality-text-global-fr-us_foreign_policy/test/28", "prediction": 0, "probabilities": [0.9986434578895569, 0.001053644111379981, 0.0003028486971743405], "tokens": 118} +{"id": "modality-text-global-de-virology/test/107", "prediction": 0, "probabilities": [0.9990602135658264, 0.00034092177520506084, 0.0005988635239191353], "tokens": 157} +{"id": "modality-text-global-ja-moral_disputes/test/331", "prediction": 0, "probabilities": [0.979249894618988, 0.0006421060534194112, 0.020107997581362724], "tokens": 88} +{"id": "modality-text-global-es-college_chemistry/test/55", "prediction": 0, "probabilities": [0.9999350309371948, 2.7789494197349995e-05, 3.717064100783318e-05], "tokens": 95} +{"id": "modality-text-global-fr-medical_genetics/test/38", "prediction": 0, "probabilities": [0.9998884201049805, 0.00010400406608823687, 7.524794909841148e-06], "tokens": 70} +{"id": "modality-text-global-de-moral_scenarios/test/833", "prediction": 0, "probabilities": [0.9981503486633301, 0.001577101182192564, 0.00027259517810307443], "tokens": 123} +{"id": "modality-text-global-fr-college_biology/test/28", "prediction": 0, "probabilities": [0.9974231719970703, 0.0025050914846360683, 7.165290298871696e-05], "tokens": 92} +{"id": "modality-text-global-de-world_religions/test/23", "prediction": 0, "probabilities": [0.9999963045120239, 3.158471599817858e-06, 5.97069117702631e-07], "tokens": 57} +{"id": "modality-text-global-fr-international_law/test/92", "prediction": 0, "probabilities": [0.9988182187080383, 0.0006443189922720194, 0.0005375103792175651], "tokens": 99} +{"id": "modality-text-global-en-philosophy/test/52", "prediction": 0, "probabilities": [0.9918150305747986, 0.00582974823191762, 0.0023552009370177984], "tokens": 53} +{"id": "modality-text-global-zh-high_school_world_history/test/209", "prediction": 2, "probabilities": [0.1980181634426117, 0.0025598679203540087, 0.7994219660758972], "tokens": 473} +{"id": "modality-text-global-fr-philosophy/test/165", "prediction": 0, "probabilities": [0.9995697140693665, 0.0004015291924588382, 2.866500653908588e-05], "tokens": 70} +{"id": "modality-text-global-zh-moral_disputes/test/36", "prediction": 0, "probabilities": [0.9999986886978149, 5.645914029628329e-07, 7.296551416402508e-07], "tokens": 49} +{"id": "modality-text-global-ja-high_school_physics/test/133", "prediction": 0, "probabilities": [0.9999765157699585, 4.76300755281045e-07, 2.300518099218607e-05], "tokens": 69} +{"id": "modality-text-global-de-high_school_statistics/test/145", "prediction": 0, "probabilities": [0.9999853372573853, 6.721969043610443e-07, 1.3990726984047797e-05], "tokens": 171} +{"id": "modality-text-global-de-high_school_psychology/test/436", "prediction": 0, "probabilities": [0.999601423740387, 3.8050886359997094e-05, 0.000360638223355636], "tokens": 54} +{"id": "modality-text-global-es-moral_scenarios/test/833", "prediction": 0, "probabilities": [0.9965351819992065, 0.0033440380357205868, 0.00012079990119673312], "tokens": 142} +{"id": "modality-text-global-en-security_studies/test/107", "prediction": 0, "probabilities": [0.6511353254318237, 0.2835375964641571, 0.06532708555459976], "tokens": 153} +{"id": "modality-text-global-ja-moral_scenarios/test/848", "prediction": 0, "probabilities": [0.9888872504234314, 0.0003878952411469072, 0.010724927298724651], "tokens": 117} +{"id": "modality-text-global-de-high_school_psychology/test/137", "prediction": 0, "probabilities": [0.9975618124008179, 0.0011778556508943439, 0.001260261400602758], "tokens": 108} +{"id": "modality-text-global-zh-professional_law/test/789", "prediction": 0, "probabilities": [0.6356386542320251, 0.00013921409845352173, 0.36422213912010193], "tokens": 384} +{"id": "modality-text-global-en-nutrition/test/129", "prediction": 0, "probabilities": [0.9941272735595703, 0.004543730523437262, 0.0013290166389197111], "tokens": 100} +{"id": "modality-text-global-fr-professional_law/test/378", "prediction": 0, "probabilities": [0.9989585876464844, 0.0008433384355157614, 0.00019807371427305043], "tokens": 293} +{"id": "modality-text-global-ja-miscellaneous/test/728", "prediction": 0, "probabilities": [0.9999171495437622, 6.156240124255419e-05, 2.1281952285789885e-05], "tokens": 39} +{"id": "modality-text-global-de-college_mathematics/test/53", "prediction": 0, "probabilities": [0.9999973773956299, 3.897895908266946e-07, 2.2610652194998693e-06], "tokens": 57} +{"id": "modality-text-global-de-jurisprudence/test/41", "prediction": 0, "probabilities": [0.9995672106742859, 0.00017442370881326497, 0.0002583370078355074], "tokens": 112} +{"id": "modality-text-global-zh-logical_fallacies/test/152", "prediction": 0, "probabilities": [0.9982739686965942, 0.0009764726855792105, 0.0007496470934711397], "tokens": 40} +{"id": "modality-text-global-en-moral_scenarios/test/56", "prediction": 0, "probabilities": [0.995468258857727, 0.0038578214589506388, 0.0006739340024068952], "tokens": 100} +{"id": "modality-text-global-zh-professional_law/test/294", "prediction": 0, "probabilities": [0.9773977398872375, 0.0007901861099526286, 0.021812060847878456], "tokens": 80} +{"id": "modality-text-global-ja-security_studies/test/19", "prediction": 2, "probabilities": [0.41343608498573303, 0.001069961697794497, 0.5854939222335815], "tokens": 374} +{"id": "modality-text-global-de-nutrition/test/301", "prediction": 0, "probabilities": [0.9999963045120239, 2.2884594272909453e-06, 1.478534272791876e-06], "tokens": 94} +{"id": "modality-text-global-fr-marketing/test/84", "prediction": 0, "probabilities": [0.995557963848114, 0.004328498151153326, 0.0001135563725256361], "tokens": 64} +{"id": "modality-text-global-ja-professional_law/test/829", "prediction": 0, "probabilities": [0.9215167164802551, 0.0001420643093297258, 0.07834126055240631], "tokens": 216} +{"id": "modality-text-global-es-moral_scenarios/test/51", "prediction": 0, "probabilities": [0.993729829788208, 0.005973490886390209, 0.0002967215550597757], "tokens": 148} +{"id": "modality-text-global-es-virology/test/64", "prediction": 0, "probabilities": [0.9999114274978638, 6.489814404631034e-05, 2.3711365429335274e-05], "tokens": 72} +{"id": "modality-text-global-es-human_sexuality/test/1", "prediction": 0, "probabilities": [0.9998530149459839, 4.5018878154223785e-05, 0.00010188462329097092], "tokens": 61} +{"id": "modality-text-global-de-virology/test/119", "prediction": 0, "probabilities": [0.999411940574646, 6.258836947381496e-05, 0.0005255379364825785], "tokens": 43} +{"id": "modality-text-global-de-miscellaneous/test/108", "prediction": 0, "probabilities": [0.9999743700027466, 1.8983842892339453e-05, 6.68411257720436e-06], "tokens": 36} +{"id": "modality-text-global-zh-moral_scenarios/test/628", "prediction": 0, "probabilities": [0.9971170425415039, 0.0002602600143291056, 0.0026227037888020277], "tokens": 104} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/24", "prediction": 0, "probabilities": [0.9989042282104492, 8.800346404314041e-05, 0.0010077361948788166], "tokens": 44} +{"id": "modality-text-global-zh-high_school_psychology/test/137", "prediction": 0, "probabilities": [0.9362800717353821, 0.004834610037505627, 0.05888529121875763], "tokens": 113} +{"id": "modality-text-global-de-professional_law/test/1383", "prediction": 0, "probabilities": [0.9999148845672607, 1.4282408301369287e-05, 7.076487236190587e-05], "tokens": 165} +{"id": "modality-text-global-zh-international_law/test/17", "prediction": 0, "probabilities": [0.9991376399993896, 1.6363710528821684e-05, 0.000846001086756587], "tokens": 63} +{"id": "modality-text-global-ja-marketing/test/160", "prediction": 0, "probabilities": [0.9987976551055908, 5.038164545112522e-06, 0.0011973854852840304], "tokens": 69} +{"id": "modality-text-global-es-high_school_psychology/test/250", "prediction": 0, "probabilities": [0.9952129125595093, 0.0025770184583961964, 0.0022100459318608046], "tokens": 66} +{"id": "modality-text-global-de-computer_security/test/86", "prediction": 0, "probabilities": [0.9998486042022705, 4.911690120934509e-05, 0.00010230702901026234], "tokens": 99} +{"id": "modality-text-global-en-miscellaneous/test/251", "prediction": 0, "probabilities": [0.9985981583595276, 0.001286679762415588, 0.00011510828335303813], "tokens": 29} +{"id": "modality-text-global-en-moral_scenarios/test/477", "prediction": 0, "probabilities": [0.9723906517028809, 0.02501102350652218, 0.002598300576210022], "tokens": 114} +{"id": "modality-text-global-fr-moral_disputes/test/224", "prediction": 0, "probabilities": [0.9998819828033447, 0.00011517924576764926, 2.841650484697311e-06], "tokens": 55} +{"id": "modality-text-global-zh-professional_accounting/test/54", "prediction": 0, "probabilities": [0.9513466954231262, 0.00014319439651444554, 0.04851005971431732], "tokens": 83} +{"id": "modality-text-global-fr-professional_law/test/1020", "prediction": 0, "probabilities": [0.9471778869628906, 0.005987904034554958, 0.046834226697683334], "tokens": 316} +{"id": "modality-text-global-en-high_school_chemistry/test/88", "prediction": 0, "probabilities": [0.9997701048851013, 0.0001429519907105714, 8.704345964360982e-05], "tokens": 82} +{"id": "modality-text-global-fr-moral_scenarios/test/94", "prediction": 0, "probabilities": [0.9998412132263184, 0.00015483966853935272, 3.894279871019535e-06], "tokens": 167} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/191", "prediction": 0, "probabilities": [0.9998643398284912, 3.205697066732682e-05, 0.00010359747830079868], "tokens": 48} +{"id": "modality-text-global-de-high_school_psychology/test/320", "prediction": 0, "probabilities": [0.9999725818634033, 1.57707017933717e-05, 1.1651077329588588e-05], "tokens": 43} +{"id": "modality-text-global-en-elementary_mathematics/test/315", "prediction": 0, "probabilities": [0.9999992847442627, 5.876656814507442e-07, 1.157484703639966e-07], "tokens": 44} +{"id": "modality-text-global-de-virology/test/118", "prediction": 0, "probabilities": [0.9999582767486572, 1.7073674825951457e-05, 2.4652419597259723e-05], "tokens": 68} +{"id": "modality-text-global-fr-nutrition/test/92", "prediction": 0, "probabilities": [0.9999487400054932, 4.7739988076500595e-05, 3.5204802770749666e-06], "tokens": 65} +{"id": "modality-text-global-zh-computer_security/test/41", "prediction": 0, "probabilities": [0.9830145835876465, 0.0013638111995533109, 0.015621588565409184], "tokens": 51} +{"id": "modality-text-global-en-professional_law/test/1072", "prediction": 0, "probabilities": [0.5645438432693481, 0.33104246854782104, 0.10441369563341141], "tokens": 326} +{"id": "modality-text-global-fr-business_ethics/test/56", "prediction": 0, "probabilities": [0.9817626476287842, 0.017786258831620216, 0.00045100002898834646], "tokens": 113} +{"id": "modality-text-global-en-sociology/test/129", "prediction": 0, "probabilities": [0.9926530122756958, 0.006006724666804075, 0.001340261078439653], "tokens": 67} +{"id": "modality-text-global-de-miscellaneous/test/413", "prediction": 0, "probabilities": [0.9998279809951782, 0.00011317827738821507, 5.8741599787026644e-05], "tokens": 53} +{"id": "modality-text-global-es-world_religions/test/37", "prediction": 0, "probabilities": [0.9992998838424683, 0.00044858604087494314, 0.0002515149535611272], "tokens": 44} +{"id": "modality-text-global-en-public_relations/test/31", "prediction": 0, "probabilities": [0.9452663064002991, 0.04299790412187576, 0.011735762469470501], "tokens": 52} +{"id": "modality-text-global-fr-philosophy/test/6", "prediction": 0, "probabilities": [0.9911304712295532, 0.00816250964999199, 0.000707023311406374], "tokens": 71} +{"id": "modality-text-global-en-moral_scenarios/test/630", "prediction": 0, "probabilities": [0.9889916181564331, 0.009282417595386505, 0.00172607006970793], "tokens": 120} +{"id": "modality-text-global-es-high_school_government_and_politics/test/135", "prediction": 0, "probabilities": [0.9946740865707397, 0.003788477974012494, 0.0015374079812318087], "tokens": 52} +{"id": "modality-text-global-zh-moral_scenarios/test/240", "prediction": 0, "probabilities": [0.9910780787467957, 0.0009736292995512486, 0.007948307320475578], "tokens": 112} +{"id": "modality-text-global-fr-us_foreign_policy/test/38", "prediction": 0, "probabilities": [0.9995712637901306, 0.0004017569590359926, 2.6903040634351783e-05], "tokens": 107} +{"id": "modality-text-global-zh-professional_psychology/test/412", "prediction": 0, "probabilities": [0.8772127032279968, 0.0006498057045973837, 0.12213741987943649], "tokens": 58} +{"id": "modality-text-global-fr-moral_scenarios/test/843", "prediction": 0, "probabilities": [0.999845027923584, 0.00015173101564869285, 3.2126906717167003e-06], "tokens": 157} +{"id": "modality-text-global-ja-professional_law/test/1248", "prediction": 2, "probabilities": [0.3059968948364258, 0.0012545465724542737, 0.692748486995697], "tokens": 401} +{"id": "modality-text-global-en-high_school_geography/test/44", "prediction": 0, "probabilities": [0.6338768005371094, 0.28810763359069824, 0.07801553606987], "tokens": 51} +{"id": "modality-text-global-fr-anatomy/test/115", "prediction": 0, "probabilities": [0.9999979734420776, 9.137398819802911e-07, 1.117052647714445e-06], "tokens": 65} +{"id": "modality-text-global-fr-security_studies/test/238", "prediction": 0, "probabilities": [0.9863191843032837, 0.010914881713688374, 0.0027658664621412754], "tokens": 348} +{"id": "modality-text-global-zh-high_school_psychology/test/320", "prediction": 0, "probabilities": [0.9957940578460693, 0.0010021814377978444, 0.003203808795660734], "tokens": 40} +{"id": "modality-text-global-zh-moral_scenarios/test/573", "prediction": 0, "probabilities": [0.9983068704605103, 0.00012768404849339277, 0.001565377926453948], "tokens": 107} +{"id": "modality-text-global-ja-clinical_knowledge/test/152", "prediction": 0, "probabilities": [0.9999799728393555, 1.7636398297327105e-06, 1.8191060007666238e-05], "tokens": 44} +{"id": "modality-text-global-es-world_religions/test/133", "prediction": 0, "probabilities": [0.9999079704284668, 5.530199996428564e-05, 3.665428812382743e-05], "tokens": 36} +{"id": "modality-text-global-de-professional_law/test/479", "prediction": 0, "probabilities": [0.9995473027229309, 7.6127703323436435e-06, 0.00044509535655379295], "tokens": 565} +{"id": "modality-text-global-en-high_school_chemistry/test/21", "prediction": 0, "probabilities": [0.9993917942047119, 0.0004462224314920604, 0.0001620408584130928], "tokens": 62} +{"id": "modality-text-global-fr-nutrition/test/145", "prediction": 0, "probabilities": [0.9976251721382141, 0.002308232942596078, 6.665564433205873e-05], "tokens": 93} +{"id": "modality-text-global-fr-human_sexuality/test/46", "prediction": 0, "probabilities": [0.9993749260902405, 0.0006213576416485012, 3.746892389244749e-06], "tokens": 61} +{"id": "modality-text-global-fr-human_aging/test/29", "prediction": 0, "probabilities": [0.9993987083435059, 0.00031541529460810125, 0.00028598494827747345], "tokens": 62} +{"id": "modality-text-global-en-professional_law/test/479", "prediction": 0, "probabilities": [0.7552739381790161, 0.1447623372077942, 0.09996373951435089], "tokens": 379} +{"id": "modality-text-global-es-high_school_microeconomics/test/123", "prediction": 0, "probabilities": [0.9999877214431763, 4.384614385344321e-06, 7.899428965174593e-06], "tokens": 68} +{"id": "modality-text-global-ja-miscellaneous/test/624", "prediction": 0, "probabilities": [0.9992539286613464, 2.604035944386851e-05, 0.0007199806277640164], "tokens": 44} +{"id": "modality-text-global-fr-moral_scenarios/test/272", "prediction": 0, "probabilities": [0.9992457628250122, 0.00074840645538643, 5.872533620276954e-06], "tokens": 153} +{"id": "modality-text-global-ja-professional_law/test/789", "prediction": 0, "probabilities": [0.694538950920105, 0.0004421313642524183, 0.30501890182495117], "tokens": 459} +{"id": "modality-text-global-en-professional_law/test/1020", "prediction": 2, "probabilities": [0.18600617349147797, 0.3469487726688385, 0.4670450985431671], "tokens": 227} +{"id": "modality-text-global-ja-miscellaneous/test/526", "prediction": 0, "probabilities": [0.9999983310699463, 4.4961673495436116e-08, 1.7273277990170754e-06], "tokens": 42} +{"id": "modality-text-global-en-moral_scenarios/test/622", "prediction": 0, "probabilities": [0.9876036643981934, 0.010516049340367317, 0.0018803034909069538], "tokens": 113} +{"id": "modality-text-global-en-virology/test/131", "prediction": 0, "probabilities": [0.9998972415924072, 6.904302426846698e-05, 3.37049896188546e-05], "tokens": 43} +{"id": "modality-text-global-es-clinical_knowledge/test/193", "prediction": 0, "probabilities": [0.9733284711837769, 0.014908826909959316, 0.011762715876102448], "tokens": 63} +{"id": "modality-text-global-ja-clinical_knowledge/test/32", "prediction": 0, "probabilities": [0.9900650382041931, 0.00092076207511127, 0.009014278650283813], "tokens": 75} +{"id": "modality-text-global-zh-moral_scenarios/test/62", "prediction": 0, "probabilities": [0.9966423511505127, 0.00040183914825320244, 0.002955799223855138], "tokens": 112} +{"id": "modality-text-global-es-high_school_psychology/test/265", "prediction": 0, "probabilities": [0.9994753003120422, 0.00020401662914082408, 0.0003207138506695628], "tokens": 63} +{"id": "modality-text-global-fr-electrical_engineering/test/90", "prediction": 0, "probabilities": [0.9989650249481201, 0.0006950417882762849, 0.0003400173445697874], "tokens": 57} +{"id": "modality-text-global-ja-moral_scenarios/test/180", "prediction": 0, "probabilities": [0.9918897747993469, 0.0006344918510876596, 0.007475745864212513], "tokens": 115} +{"id": "modality-text-global-es-high_school_government_and_politics/test/100", "prediction": 0, "probabilities": [0.9941668510437012, 0.002878471976146102, 0.0029546162113547325], "tokens": 140} +{"id": "modality-text-global-ja-high_school_microeconomics/test/110", "prediction": 0, "probabilities": [0.9997829794883728, 1.0713064511946868e-05, 0.00020630711514968425], "tokens": 63} +{"id": "modality-text-global-ja-professional_accounting/test/267", "prediction": 0, "probabilities": [0.9730340242385864, 6.112065602792427e-05, 0.026904821395874023], "tokens": 122} +{"id": "modality-text-global-zh-high_school_biology/test/259", "prediction": 2, "probabilities": [0.09733547270298004, 0.0015013257507234812, 0.9011632800102234], "tokens": 154} +{"id": "modality-text-global-es-moral_scenarios/test/730", "prediction": 0, "probabilities": [0.9988046884536743, 0.0010778851574286819, 0.0001173980490420945], "tokens": 143} +{"id": "modality-text-global-fr-public_relations/test/50", "prediction": 0, "probabilities": [0.9963380098342896, 0.002383413026109338, 0.0012785530416294932], "tokens": 80} +{"id": "modality-text-global-en-professional_medicine/test/69", "prediction": 0, "probabilities": [0.7314008474349976, 0.22620543837547302, 0.04239378124475479], "tokens": 92} +{"id": "modality-text-global-es-jurisprudence/test/53", "prediction": 0, "probabilities": [0.9972831010818481, 0.0018376861698925495, 0.0008791621658019722], "tokens": 96} +{"id": "modality-text-global-fr-professional_accounting/test/271", "prediction": 0, "probabilities": [0.9940949082374573, 0.005850411020219326, 5.475763464346528e-05], "tokens": 328} +{"id": "modality-text-global-de-college_biology/test/137", "prediction": 0, "probabilities": [0.9999443292617798, 5.0018839829135686e-05, 5.545830845221644e-06], "tokens": 58} +{"id": "modality-text-global-fr-high_school_computer_science/test/19", "prediction": 0, "probabilities": [0.997272789478302, 0.0003538151504471898, 0.0023733668494969606], "tokens": 141} +{"id": "modality-text-global-en-college_medicine/test/10", "prediction": 0, "probabilities": [0.9967813491821289, 0.0015589040704071522, 0.0016597169451415539], "tokens": 75} +{"id": "modality-text-global-es-human_sexuality/test/70", "prediction": 0, "probabilities": [0.9994682669639587, 9.400584531249478e-05, 0.00043773098150268197], "tokens": 125} +{"id": "modality-text-global-ja-moral_disputes/test/196", "prediction": 0, "probabilities": [0.9983335137367249, 0.0004984153201803565, 0.0011680350871756673], "tokens": 64} +{"id": "modality-text-global-fr-philosophy/test/168", "prediction": 0, "probabilities": [0.9933573603630066, 0.0057021575048565865, 0.0009404617012478411], "tokens": 62} +{"id": "modality-text-global-en-high_school_psychology/test/265", "prediction": 0, "probabilities": [0.8931048512458801, 0.0751371756196022, 0.03175800293684006], "tokens": 45} +{"id": "modality-text-global-de-moral_scenarios/test/767", "prediction": 0, "probabilities": [0.9990735054016113, 0.0007091122679412365, 0.00021731553715653718], "tokens": 112} +{"id": "modality-text-global-zh-medical_genetics/test/49", "prediction": 0, "probabilities": [0.999992847442627, 8.619168738732697e-07, 6.314684014796512e-06], "tokens": 30} +{"id": "modality-text-global-es-high_school_geography/test/142", "prediction": 0, "probabilities": [0.9994003772735596, 0.0005339908530004323, 6.568118260474876e-05], "tokens": 52} +{"id": "modality-text-global-fr-sociology/test/71", "prediction": 0, "probabilities": [0.997755229473114, 0.0013201169203966856, 0.0009245952242054045], "tokens": 160} +{"id": "modality-text-global-es-moral_scenarios/test/480", "prediction": 0, "probabilities": [0.9952994585037231, 0.004564563278108835, 0.00013599421072285622], "tokens": 134} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/100", "prediction": 0, "probabilities": [0.6326652765274048, 0.0004335654084570706, 0.366901159286499], "tokens": 144} +{"id": "modality-text-global-fr-jurisprudence/test/42", "prediction": 0, "probabilities": [0.996791660785675, 0.0026393565349280834, 0.0005689720273949206], "tokens": 38} +{"id": "modality-text-global-ja-clinical_knowledge/test/84", "prediction": 0, "probabilities": [0.9676151275634766, 1.2934722690260969e-05, 0.032371893525123596], "tokens": 86} +{"id": "modality-text-global-de-high_school_mathematics/test/45", "prediction": 0, "probabilities": [0.9999982118606567, 4.941952624903934e-09, 1.814383836062916e-06], "tokens": 88} +{"id": "modality-text-global-en-professional_law/test/1383", "prediction": 0, "probabilities": [0.8906983733177185, 0.04645107686519623, 0.06285055726766586], "tokens": 136} +{"id": "modality-text-global-fr-philosophy/test/139", "prediction": 0, "probabilities": [0.9974203109741211, 0.000868347764480859, 0.0017113268841058016], "tokens": 77} +{"id": "modality-text-global-en-professional_law/test/1189", "prediction": 0, "probabilities": [0.6837632060050964, 0.23161494731903076, 0.0846218690276146], "tokens": 235} +{"id": "modality-text-global-en-marketing/test/90", "prediction": 0, "probabilities": [0.9378598928451538, 0.05344052240252495, 0.008699574507772923], "tokens": 59} +{"id": "modality-text-global-zh-sociology/test/14", "prediction": 0, "probabilities": [0.9460058212280273, 0.0070934840478003025, 0.04690070450305939], "tokens": 103} +{"id": "modality-text-global-zh-high_school_government_and_politics/test/58", "prediction": 0, "probabilities": [0.9915021657943726, 0.00032003512023948133, 0.008177789859473705], "tokens": 78} +{"id": "modality-text-global-ja-nutrition/test/301", "prediction": 0, "probabilities": [0.9982931017875671, 6.1848986661061645e-06, 0.0017007037531584501], "tokens": 114} +{"id": "modality-text-global-fr-high_school_psychology/test/83", "prediction": 0, "probabilities": [0.99272221326828, 0.007044618483632803, 0.00023308278468903154], "tokens": 60} +{"id": "modality-text-global-de-moral_disputes/test/200", "prediction": 0, "probabilities": [0.9997941851615906, 9.470537042943761e-05, 0.0001110927332774736], "tokens": 102} +{"id": "modality-text-global-en-philosophy/test/104", "prediction": 0, "probabilities": [0.9142885804176331, 0.0742788016796112, 0.011432548053562641], "tokens": 74} +{"id": "modality-text-global-en-professional_law/test/774", "prediction": 0, "probabilities": [0.6293821334838867, 0.22553063929080963, 0.14508722722530365], "tokens": 359} +{"id": "modality-text-global-zh-virology/test/85", "prediction": 0, "probabilities": [0.9693482518196106, 0.0012213598238304257, 0.02943047136068344], "tokens": 52} +{"id": "modality-text-global-en-professional_law/test/261", "prediction": 0, "probabilities": [0.9218918085098267, 0.052891068160533905, 0.02521713450551033], "tokens": 146} +{"id": "modality-text-global-zh-high_school_psychology/test/456", "prediction": 0, "probabilities": [0.8978177309036255, 0.0004935595788992941, 0.10168877989053726], "tokens": 66} +{"id": "modality-text-global-es-moral_scenarios/test/153", "prediction": 0, "probabilities": [0.9977756142616272, 0.0021025845780968666, 0.00012181521742604673], "tokens": 172} +{"id": "modality-text-global-en-professional_psychology/test/128", "prediction": 0, "probabilities": [0.9681249260902405, 0.01272379606962204, 0.019151240587234497], "tokens": 64} +{"id": "modality-text-global-fr-miscellaneous/test/58", "prediction": 0, "probabilities": [0.9808623790740967, 0.018163684755563736, 0.0009738916414789855], "tokens": 35} +{"id": "modality-text-global-es-moral_scenarios/test/786", "prediction": 0, "probabilities": [0.9981860518455505, 0.0017238055588677526, 9.005706669995561e-05], "tokens": 137} +{"id": "modality-text-global-es-moral_scenarios/test/249", "prediction": 0, "probabilities": [0.9983164072036743, 0.0015766589203849435, 0.00010699915583245456], "tokens": 146} +{"id": "modality-text-global-ja-virology/test/107", "prediction": 0, "probabilities": [0.9300639629364014, 0.0008466494618915021, 0.06908939778804779], "tokens": 146} +{"id": "modality-text-global-de-high_school_biology/test/47", "prediction": 0, "probabilities": [0.9999722242355347, 1.4555106645275373e-05, 1.3179298548493534e-05], "tokens": 127} +{"id": "modality-text-global-es-human_sexuality/test/41", "prediction": 0, "probabilities": [0.9999980926513672, 1.0697607422116562e-06, 8.778691267252725e-07], "tokens": 46} +{"id": "modality-text-global-ja-miscellaneous/test/423", "prediction": 0, "probabilities": [0.9640770554542542, 0.00024947492056526244, 0.03567342087626457], "tokens": 54} +{"id": "modality-text-global-es-high_school_government_and_politics/test/114", "prediction": 0, "probabilities": [0.9944211840629578, 0.005383000243455172, 0.00019578184583224356], "tokens": 70} +{"id": "modality-text-global-en-international_law/test/33", "prediction": 0, "probabilities": [0.7790461182594299, 0.019402507692575455, 0.20155136287212372], "tokens": 107} +{"id": "modality-text-global-zh-moral_scenarios/test/155", "prediction": 0, "probabilities": [0.996609091758728, 0.00014209722576197237, 0.003248789580538869], "tokens": 104} +{"id": "modality-text-global-zh-moral_scenarios/test/274", "prediction": 0, "probabilities": [0.9794308543205261, 0.0006764094578102231, 0.019892709329724312], "tokens": 128} +{"id": "modality-text-global-de-professional_psychology/test/512", "prediction": 0, "probabilities": [0.9999420642852783, 1.2334281564108096e-05, 4.571362296701409e-05], "tokens": 54} +{"id": "modality-text-global-fr-moral_scenarios/test/240", "prediction": 0, "probabilities": [0.9979432225227356, 0.002048199065029621, 8.54885729495436e-06], "tokens": 170} +{"id": "modality-text-global-zh-moral_scenarios/test/281", "prediction": 0, "probabilities": [0.9983036518096924, 0.0002473288623150438, 0.0014490772737190127], "tokens": 112} +{"id": "modality-text-global-es-professional_psychology/test/80", "prediction": 0, "probabilities": [0.9962892532348633, 0.0030037795659154654, 0.0007070358842611313], "tokens": 46} +{"id": "modality-text-global-ja-marketing/test/58", "prediction": 0, "probabilities": [0.9986447691917419, 7.956902118166909e-05, 0.0012756893411278725], "tokens": 43} +{"id": "modality-text-global-fr-college_physics/test/77", "prediction": 0, "probabilities": [0.9986470341682434, 0.0012481395388022065, 0.00010491882130736485], "tokens": 119} +{"id": "modality-text-global-es-high_school_psychology/test/367", "prediction": 0, "probabilities": [0.9999998807907104, 4.4045371794254606e-08, 1.0357342716815765e-07], "tokens": 38} +{"id": "modality-text-global-es-high_school_microeconomics/test/153", "prediction": 0, "probabilities": [0.99892657995224, 0.001032795524224639, 4.065332541358657e-05], "tokens": 56} +{"id": "modality-text-global-en-elementary_mathematics/test/366", "prediction": 0, "probabilities": [0.9999597072601318, 3.616129470174201e-05, 4.184022145636845e-06], "tokens": 48} +{"id": "modality-text-global-es-virology/test/163", "prediction": 0, "probabilities": [0.9992709755897522, 0.0006821799324825406, 4.681389327743091e-05], "tokens": 51} +{"id": "modality-text-global-de-high_school_statistics/test/198", "prediction": 0, "probabilities": [0.999983549118042, 2.5359063329233322e-06, 1.3997468158777338e-05], "tokens": 103} +{"id": "modality-text-global-de-moral_disputes/test/36", "prediction": 0, "probabilities": [0.999990701675415, 8.336589417012874e-06, 9.201868920172274e-07], "tokens": 58} +{"id": "modality-text-global-en-us_foreign_policy/test/77", "prediction": 0, "probabilities": [0.920806884765625, 0.0626310333609581, 0.01656203158199787], "tokens": 90} +{"id": "modality-text-global-ja-philosophy/test/39", "prediction": 0, "probabilities": [0.9083856344223022, 0.004331659059971571, 0.0872826874256134], "tokens": 72} +{"id": "modality-text-global-fr-philosophy/test/142", "prediction": 0, "probabilities": [0.9993370175361633, 0.0006402151193469763, 2.2749725758330896e-05], "tokens": 53} +{"id": "modality-text-global-ja-high_school_chemistry/test/33", "prediction": 0, "probabilities": [0.9969441294670105, 5.7302662753500044e-05, 0.0029986232984811068], "tokens": 68} +{"id": "modality-text-global-de-high_school_biology/test/166", "prediction": 0, "probabilities": [0.9989114999771118, 0.00026330864056944847, 0.0008252265979535878], "tokens": 83} +{"id": "modality-text-global-zh-public_relations/test/66", "prediction": 0, "probabilities": [0.9701888561248779, 0.0002648538211360574, 0.02954627014696598], "tokens": 67} +{"id": "modality-text-global-ja-moral_scenarios/test/295", "prediction": 0, "probabilities": [0.9879366755485535, 0.0008808263228274882, 0.011182469315826893], "tokens": 126} +{"id": "modality-text-global-zh-professional_psychology/test/472", "prediction": 0, "probabilities": [0.9867091178894043, 0.0003236318880226463, 0.012967274524271488], "tokens": 72} +{"id": "modality-text-global-ja-miscellaneous/test/14", "prediction": 2, "probabilities": [0.09889689832925797, 0.002613396616652608, 0.8984897136688232], "tokens": 101} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/274", "prediction": 0, "probabilities": [0.9997153878211975, 6.580130047950661e-06, 0.00027798116207122803], "tokens": 48} +{"id": "modality-text-global-de-miscellaneous/test/728", "prediction": 0, "probabilities": [0.999997615814209, 1.545062900731864e-06, 7.989701202859578e-07], "tokens": 29} +{"id": "modality-text-global-es-clinical_knowledge/test/80", "prediction": 0, "probabilities": [0.9988536834716797, 5.472190969157964e-05, 0.0010915134334936738], "tokens": 56} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/110", "prediction": 0, "probabilities": [0.9998934268951416, 2.5701763206598116e-06, 0.0001039969502016902], "tokens": 73} +{"id": "modality-text-global-en-college_medicine/test/73", "prediction": 0, "probabilities": [0.9982172846794128, 0.0009256203775294125, 0.0008570693898946047], "tokens": 89} +{"id": "modality-text-global-zh-high_school_biology/test/96", "prediction": 0, "probabilities": [0.9805221557617188, 0.000989061314612627, 0.018488747999072075], "tokens": 78} +{"id": "modality-text-global-fr-sociology/test/188", "prediction": 0, "probabilities": [0.9992407560348511, 0.0007258023251779377, 3.343878415762447e-05], "tokens": 79} +{"id": "modality-text-global-de-moral_scenarios/test/587", "prediction": 0, "probabilities": [0.9986907839775085, 0.0009732343023642898, 0.0003359777037985623], "tokens": 127} +{"id": "modality-text-global-zh-high_school_microeconomics/test/176", "prediction": 0, "probabilities": [0.9713844656944275, 0.01255414541810751, 0.016061443835496902], "tokens": 78} +{"id": "modality-text-global-en-high_school_macroeconomics/test/354", "prediction": 0, "probabilities": [0.9936447143554688, 0.0040568700060248375, 0.002298438921570778], "tokens": 71} +{"id": "modality-text-global-de-miscellaneous/test/365", "prediction": 0, "probabilities": [0.9978235960006714, 0.0020842202939093113, 9.224232053384185e-05], "tokens": 42} +{"id": "modality-text-global-ja-philosophy/test/101", "prediction": 0, "probabilities": [0.9824246764183044, 7.356939022429287e-05, 0.017501775175333023], "tokens": 101} +{"id": "modality-text-global-de-moral_scenarios/test/111", "prediction": 0, "probabilities": [0.9993821382522583, 0.0004116925410926342, 0.0002061086124740541], "tokens": 129} +{"id": "modality-text-global-ja-professional_medicine/test/127", "prediction": 0, "probabilities": [0.5231147408485413, 0.00014241255121305585, 0.4767428934574127], "tokens": 191} +{"id": "modality-text-global-zh-conceptual_physics/test/10", "prediction": 0, "probabilities": [0.9789097905158997, 0.0010200005490332842, 0.020070107653737068], "tokens": 51} +{"id": "modality-text-global-fr-moral_scenarios/test/148", "prediction": 0, "probabilities": [0.9992764592170715, 0.0007069396669976413, 1.6534140740986913e-05], "tokens": 174} +{"id": "modality-text-global-ja-high_school_chemistry/test/88", "prediction": 0, "probabilities": [0.9990829229354858, 2.151152875740081e-05, 0.0008954883669503033], "tokens": 86} +{"id": "modality-text-global-en-high_school_government_and_politics/test/117", "prediction": 0, "probabilities": [0.9983664155006409, 0.0012322813272476196, 0.0004013525031041354], "tokens": 55} +{"id": "modality-text-global-zh-elementary_mathematics/test/325", "prediction": 0, "probabilities": [0.9960201382637024, 7.440214540110901e-05, 0.0039054546505212784], "tokens": 36} +{"id": "modality-text-global-de-professional_psychology/test/279", "prediction": 0, "probabilities": [0.9992513060569763, 0.0006575288134627044, 9.109728853218257e-05], "tokens": 36} +{"id": "modality-text-global-ja-computer_security/test/86", "prediction": 0, "probabilities": [0.9948040843009949, 0.00018054453539662063, 0.005015425384044647], "tokens": 94} +{"id": "modality-text-global-de-international_law/test/8", "prediction": 0, "probabilities": [0.9999634027481079, 1.4691143405798357e-05, 2.1978181393933482e-05], "tokens": 97} +{"id": "modality-text-global-zh-high_school_psychology/test/510", "prediction": 0, "probabilities": [0.9974690675735474, 0.0006164445076137781, 0.0019144938560202718], "tokens": 39} +{"id": "modality-text-global-ja-moral_disputes/test/232", "prediction": 0, "probabilities": [0.9993602633476257, 5.5932887335075065e-05, 0.0005838401848450303], "tokens": 41} +{"id": "modality-text-global-ja-nutrition/test/297", "prediction": 0, "probabilities": [0.9975814819335938, 1.757877726049628e-05, 0.0024008764885365963], "tokens": 136} +{"id": "modality-text-global-zh-high_school_geography/test/110", "prediction": 0, "probabilities": [0.9926550388336182, 0.0005986044416204095, 0.006746285129338503], "tokens": 50} +{"id": "modality-text-global-ja-nutrition/test/129", "prediction": 0, "probabilities": [0.9983057975769043, 9.667088306741789e-06, 0.0016845888458192348], "tokens": 127} +{"id": "modality-text-global-ja-moral_scenarios/test/456", "prediction": 0, "probabilities": [0.9896334409713745, 0.0004580530512612313, 0.009908520616590977], "tokens": 120} +{"id": "modality-text-global-zh-professional_law/test/500", "prediction": 2, "probabilities": [0.4978630840778351, 0.000740795221645385, 0.501396119594574], "tokens": 173} +{"id": "modality-text-global-zh-human_aging/test/127", "prediction": 0, "probabilities": [0.9979725480079651, 4.888640614808537e-05, 0.001978551037609577], "tokens": 56} +{"id": "modality-text-global-en-high_school_government_and_politics/test/114", "prediction": 0, "probabilities": [0.9712136387825012, 0.0274069644510746, 0.0013793535763397813], "tokens": 59} +{"id": "modality-text-global-es-high_school_psychology/test/273", "prediction": 0, "probabilities": [0.9977079629898071, 0.001710068783722818, 0.0005819520447403193], "tokens": 72} +{"id": "modality-text-global-de-high_school_chemistry/test/124", "prediction": 0, "probabilities": [0.999675989151001, 0.00016346214397344738, 0.0001605740690138191], "tokens": 47} +{"id": "modality-text-global-en-moral_disputes/test/273", "prediction": 0, "probabilities": [0.9923514723777771, 0.0054028769955039024, 0.002245632465928793], "tokens": 53} +{"id": "modality-text-global-de-high_school_macroeconomics/test/284", "prediction": 0, "probabilities": [0.9999570846557617, 1.5534107660641894e-05, 2.747367943811696e-05], "tokens": 112} +{"id": "modality-text-global-de-high_school_world_history/test/10", "prediction": 0, "probabilities": [0.9996745586395264, 0.00015697655908297747, 0.00016856456932146102], "tokens": 321} +{"id": "modality-text-global-es-conceptual_physics/test/164", "prediction": 0, "probabilities": [0.9998610019683838, 7.559837104054168e-05, 6.343512359308079e-05], "tokens": 34} +{"id": "modality-text-global-de-human_aging/test/53", "prediction": 0, "probabilities": [0.9970807433128357, 0.0019971018191426992, 0.0009221669170074165], "tokens": 53} +{"id": "modality-text-global-es-security_studies/test/140", "prediction": 0, "probabilities": [0.9982092380523682, 0.0014565378660336137, 0.00033431468182243407], "tokens": 135} +{"id": "modality-text-global-de-philosophy/test/202", "prediction": 0, "probabilities": [0.9958330988883972, 0.0039586653001606464, 0.0002082778955809772], "tokens": 50} +{"id": "modality-text-global-es-professional_law/test/807", "prediction": 0, "probabilities": [0.99481201171875, 0.003735407954081893, 0.001452591735869646], "tokens": 308} +{"id": "modality-text-global-es-public_relations/test/61", "prediction": 0, "probabilities": [0.9841779470443726, 0.006818933878093958, 0.009003178216516972], "tokens": 56} +{"id": "modality-aya-review-eng-26", "prediction": 0, "probabilities": [0.9991441965103149, 0.00029891496524214745, 0.0005567795597016811], "tokens": 52} +{"id": "modality-aya-review-spa-18", "prediction": 0, "probabilities": [0.9997243285179138, 0.00016004857025109231, 0.00011557234392967075], "tokens": 10} +{"id": "modality-aya-review-spa-32", "prediction": 0, "probabilities": [0.9996076226234436, 3.370043486938812e-05, 0.00035870351712219417], "tokens": 17} +{"id": "modality-aya-review-fra-0", "prediction": 0, "probabilities": [0.9964900612831116, 0.0005694783758372068, 0.0029403732623904943], "tokens": 12} +{"id": "modality-aya-review-fra-2", "prediction": 0, "probabilities": [0.9998052716255188, 0.00014953817299101502, 4.5200933527667075e-05], "tokens": 22} +{"id": "modality-aya-review-fra-5", "prediction": 0, "probabilities": [0.9985508322715759, 5.870031236554496e-05, 0.001390427234582603], "tokens": 17} +{"id": "modality-aya-review-fra-6", "prediction": 0, "probabilities": [0.9991193413734436, 0.0007729232893325388, 0.0001078403220162727], "tokens": 50} +{"id": "modality-aya-review-fra-9", "prediction": 0, "probabilities": [0.9968860745429993, 0.00015497664571739733, 0.0029589238110929728], "tokens": 12} +{"id": "modality-aya-review-fra-12", "prediction": 0, "probabilities": [0.9991464614868164, 0.0001731250376906246, 0.0006803090800531209], "tokens": 10} +{"id": "modality-aya-review-fra-16", "prediction": 0, "probabilities": [0.9973340034484863, 0.0015951392706483603, 0.001070856349542737], "tokens": 10} +{"id": "modality-aya-review-fra-17", "prediction": 0, "probabilities": [0.998738706111908, 0.0005342114018276334, 0.0007271437207236886], "tokens": 11} +{"id": "modality-aya-review-fra-21", "prediction": 0, "probabilities": [0.9993771910667419, 5.353775850380771e-05, 0.0005693358252756298], "tokens": 16} +{"id": "modality-aya-review-fra-23", "prediction": 0, "probabilities": [0.9996157884597778, 8.622224413556978e-05, 0.00029789298423565924], "tokens": 10} +{"id": "modality-aya-review-fra-24", "prediction": 0, "probabilities": [0.9962731599807739, 0.00025885310606099665, 0.0034678978845477104], "tokens": 17} +{"id": "modality-aya-review-fra-25", "prediction": 0, "probabilities": [0.994699239730835, 0.0018875559326261282, 0.0034132185392081738], "tokens": 14} +{"id": "modality-aya-review-fra-26", "prediction": 0, "probabilities": [0.9971334934234619, 0.0005774012533947825, 0.0022890279069542885], "tokens": 27} +{"id": "modality-aya-review-fra-31", "prediction": 0, "probabilities": [0.9996941089630127, 0.000237916101468727, 6.806082092225552e-05], "tokens": 56} +{"id": "modality-aya-review-fra-33", "prediction": 0, "probabilities": [0.9996111989021301, 0.0003245777334086597, 6.41805527266115e-05], "tokens": 22} +{"id": "modality-aya-review-deu-19", "prediction": 0, "probabilities": [0.999366819858551, 0.0005457834922708571, 8.743038779357448e-05], "tokens": 24} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:head", "prediction": 1, "probabilities": [0.006116995122283697, 0.972854733467102, 0.02102835476398468], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:middle", "prediction": 1, "probabilities": [0.006047919392585754, 0.9731197953224182, 0.02083226665854454], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:tail", "prediction": 1, "probabilities": [0.005523096304386854, 0.9750546813011169, 0.0194222591817379], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:head", "prediction": 1, "probabilities": [0.05243253707885742, 0.898596465587616, 0.04897107928991318], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:middle", "prediction": 1, "probabilities": [0.05228396877646446, 0.9043846726417542, 0.04333141818642616], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:tail", "prediction": 1, "probabilities": [0.05819031223654747, 0.893298327922821, 0.04851140081882477], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:head", "prediction": 1, "probabilities": [0.0933023989200592, 0.8448665142059326, 0.0618310309946537], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:middle", "prediction": 1, "probabilities": [0.08570212125778198, 0.8619400262832642, 0.052357837557792664], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:tail", "prediction": 1, "probabilities": [0.09375524520874023, 0.84223473072052, 0.06401007622480392], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:head", "prediction": 1, "probabilities": [0.01892460137605667, 0.9408096671104431, 0.0402657575905323], "tokens": 32768} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:middle", "prediction": 1, "probabilities": [0.021593313664197922, 0.9420393705368042, 0.03636731579899788], "tokens": 32768} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:tail", "prediction": 1, "probabilities": [0.021051300689578056, 0.9356150031089783, 0.04333370923995972], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:head", "prediction": 1, "probabilities": [0.005844043102115393, 0.9743524193763733, 0.019803514704108238], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:middle", "prediction": 1, "probabilities": [0.005962733644992113, 0.9738060832023621, 0.0202312171459198], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:tail", "prediction": 1, "probabilities": [0.005606106482446194, 0.9754485487937927, 0.018945319578051567], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:head", "prediction": 1, "probabilities": [0.05299021676182747, 0.8995090126991272, 0.04750071093440056], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:middle", "prediction": 1, "probabilities": [0.05489592254161835, 0.9026771187782288, 0.042426999658346176], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:tail", "prediction": 1, "probabilities": [0.06004256382584572, 0.8867747187614441, 0.05318266525864601], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:head", "prediction": 1, "probabilities": [0.09716001898050308, 0.8411611318588257, 0.061678823083639145], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:middle", "prediction": 1, "probabilities": [0.09002721309661865, 0.8639159798622131, 0.046056799590587616], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:tail", "prediction": 1, "probabilities": [0.09680497646331787, 0.8385801911354065, 0.06461479514837265], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:head", "prediction": 1, "probabilities": [0.018892664462327957, 0.9408450126647949, 0.04026234522461891], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:middle", "prediction": 1, "probabilities": [0.024098830297589302, 0.9330897331237793, 0.04281148687005043], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:tail", "prediction": 1, "probabilities": [0.022436777129769325, 0.9323981404304504, 0.04516506567597389], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:head", "prediction": 1, "probabilities": [0.0064555201679468155, 0.972822904586792, 0.020721564069390297], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:middle", "prediction": 1, "probabilities": [0.006823436822742224, 0.9726866483688354, 0.02048994041979313], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:tail", "prediction": 1, "probabilities": [0.005964409559965134, 0.9752323031425476, 0.018803248181939125], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:head", "prediction": 1, "probabilities": [0.058970361948013306, 0.8878282308578491, 0.053201381117105484], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:middle", "prediction": 1, "probabilities": [0.07305113971233368, 0.8878229260444641, 0.03912598639726639], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:tail", "prediction": 1, "probabilities": [0.06414397060871124, 0.8836209774017334, 0.052235063165426254], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:head", "prediction": 1, "probabilities": [0.10934865474700928, 0.8284882307052612, 0.06216316670179367], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:middle", "prediction": 1, "probabilities": [0.09649969637393951, 0.8530899286270142, 0.05041036754846573], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:tail", "prediction": 1, "probabilities": [0.09250806272029877, 0.8538908362388611, 0.05360110104084015], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:head", "prediction": 1, "probabilities": [0.018764005973935127, 0.9415266513824463, 0.039709314703941345], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:middle", "prediction": 1, "probabilities": [0.025768449530005455, 0.938748300075531, 0.035483282059431076], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:tail", "prediction": 1, "probabilities": [0.022751467302441597, 0.9358742833137512, 0.04137421399354935], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:head", "prediction": 0, "probabilities": [0.6308252215385437, 0.0046019842848181725, 0.36457276344299316], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:middle", "prediction": 0, "probabilities": [0.6372454762458801, 0.004589333198964596, 0.3581652045249939], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:tail", "prediction": 0, "probabilities": [0.6208454370498657, 0.005745680071413517, 0.3734089732170105], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:head", "prediction": 0, "probabilities": [0.8816884756088257, 0.005601323209702969, 0.11271021515130997], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:middle", "prediction": 0, "probabilities": [0.8970577716827393, 0.007374517619609833, 0.0955677404999733], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:tail", "prediction": 0, "probabilities": [0.8741703629493713, 0.006105699110776186, 0.11972393840551376], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:head", "prediction": 0, "probabilities": [0.8367392420768738, 0.005675621796399355, 0.15758514404296875], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:middle", "prediction": 0, "probabilities": [0.8635813593864441, 0.005608693230897188, 0.1308099925518036], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:tail", "prediction": 0, "probabilities": [0.8333358764648438, 0.005379101727157831, 0.16128502786159515], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:head", "prediction": 0, "probabilities": [0.7358731031417847, 0.006951930001378059, 0.25717493891716003], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:middle", "prediction": 0, "probabilities": [0.804124653339386, 0.007713929284363985, 0.18816140294075012], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:tail", "prediction": 0, "probabilities": [0.7342798709869385, 0.006674294359982014, 0.25904580950737], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:head", "prediction": 0, "probabilities": [0.6107318997383118, 0.003131063422188163, 0.3861370384693146], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:middle", "prediction": 0, "probabilities": [0.597490131855011, 0.0032603612635284662, 0.39924943447113037], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:tail", "prediction": 0, "probabilities": [0.6233862042427063, 0.0042242733761668205, 0.37238946557044983], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:head", "prediction": 0, "probabilities": [0.8863794207572937, 0.004681477788835764, 0.10893914103507996], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:middle", "prediction": 0, "probabilities": [0.8882835507392883, 0.005308613646775484, 0.10640779882669449], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:tail", "prediction": 0, "probabilities": [0.8667288422584534, 0.0052335322834551334, 0.12803760170936584], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:head", "prediction": 0, "probabilities": [0.841464102268219, 0.004867097828537226, 0.1536688655614853], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:middle", "prediction": 0, "probabilities": [0.8596799969673157, 0.0049356697127223015, 0.13538435101509094], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:tail", "prediction": 0, "probabilities": [0.8388751745223999, 0.005156899336725473, 0.15596795082092285], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:head", "prediction": 0, "probabilities": [0.7425686717033386, 0.006529370788484812, 0.2509019374847412], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:middle", "prediction": 0, "probabilities": [0.7668247222900391, 0.007124269846826792, 0.2260509878396988], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:tail", "prediction": 0, "probabilities": [0.7463517189025879, 0.006614087615162134, 0.2470342218875885], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:head", "prediction": 0, "probabilities": [0.5914254784584045, 0.004516827408224344, 0.4040577709674835], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:middle", "prediction": 0, "probabilities": [0.6139926910400391, 0.004932440351694822, 0.38107484579086304], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:tail", "prediction": 0, "probabilities": [0.5956955552101135, 0.005606650374829769, 0.39869776368141174], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:head", "prediction": 0, "probabilities": [0.8646814227104187, 0.005881097633391619, 0.12943749129772186], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:middle", "prediction": 0, "probabilities": [0.8748485445976257, 0.0065362947061657906, 0.11861509829759598], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:tail", "prediction": 0, "probabilities": [0.8655475378036499, 0.006161091383546591, 0.12829136848449707], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:head", "prediction": 0, "probabilities": [0.8287618160247803, 0.00578682916238904, 0.16545134782791138], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:middle", "prediction": 0, "probabilities": [0.8572750687599182, 0.006861572619527578, 0.13586337864398956], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:tail", "prediction": 0, "probabilities": [0.8335837125778198, 0.005588250700384378, 0.16082806885242462], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:head", "prediction": 0, "probabilities": [0.732306182384491, 0.007071708329021931, 0.2606221139431], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:middle", "prediction": 0, "probabilities": [0.799275279045105, 0.0073036085814237595, 0.1934211552143097], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:tail", "prediction": 0, "probabilities": [0.7283777594566345, 0.006715212482959032, 0.26490703225135803], "tokens": 32768} diff --git a/evaluation/final-candidate-fp32.json b/evaluation/final-candidate-fp32.json new file mode 100644 index 0000000000000000000000000000000000000000..443aea7b660a7426706ecafdf1e7dcb7e89426fe --- /dev/null +++ b/evaluation/final-candidate-fp32.json @@ -0,0 +1,1445 @@ +{ + "rows": 1411, + "accuracy": 0.9312544294826365, + "macro_f1": 0.9274533214132176, + "macro_f1_present_labels": 0.9274533214132176, + "per_label": { + "AR": { + "support": 643, + "tp": 632, + "fp": 73, + "fn": 11, + "precision": 0.8964539007092198, + "recall": 0.9828926905132193, + "f1": 0.9376854599406528 + }, + "DIFFUSION": { + "support": 384, + "tp": 302, + "fp": 9, + "fn": 82, + "precision": 0.9710610932475884, + "recall": 0.7864583333333334, + "f1": 0.8690647482014389 + }, + "BOTH": { + "support": 384, + "tp": 380, + "fp": 15, + "fn": 4, + "precision": 0.9620253164556962, + "recall": 0.9895833333333334, + "f1": 0.975609756097561 + } + }, + "confusion": [ + [ + 632, + 5, + 6 + ], + [ + 73, + 302, + 9 + ], + [ + 0, + 4, + 380 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 240, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 240, + "tp": 240, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 240, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.75, + "macro_f1": 0.7394880644233535, + "macro_f1_present_labels": 0.7394880644233535, + "per_label": { + "AR": { + "support": 24, + "tp": 18, + "fp": 3, + "fn": 6, + "precision": 0.8571428571428571, + "recall": 0.75, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 24, + "tp": 12, + "fp": 2, + "fn": 12, + "precision": 0.8571428571428571, + "recall": 0.5, + "f1": 0.631578947368421 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 13, + "fn": 0, + "precision": 0.6486486486486487, + "recall": 1.0, + "f1": 0.7868852459016393 + } + }, + "confusion": [ + [ + 18, + 2, + 4 + ], + [ + 3, + 12, + 9 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 1080, + "accuracy": 0.9296296296296296, + "macro_f1": 0.9290630636363364, + "macro_f1_present_labels": 0.9290630636363364, + "per_label": { + "AR": { + "support": 360, + "tp": 358, + "fp": 70, + "fn": 2, + "precision": 0.8364485981308412, + "recall": 0.9944444444444445, + "f1": 0.9086294416243654 + }, + "DIFFUSION": { + "support": 360, + "tp": 290, + "fp": 6, + "fn": 70, + "precision": 0.9797297297297297, + "recall": 0.8055555555555556, + "f1": 0.8841463414634146 + }, + "BOTH": { + "support": 360, + "tp": 356, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.9888888888888889, + "f1": 0.994413407821229 + } + }, + "confusion": [ + [ + 358, + 2, + 0 + ], + [ + 70, + 290, + 0 + ], + [ + 0, + 4, + 356 + ] + ] + }, + "aya-original-human": { + "rows": 19, + "accuracy": 0.8421052631578947, + "macro_f1": 0.30476190476190473, + "macro_f1_present_labels": 0.9142857142857143, + "per_label": { + "AR": { + "support": 19, + "tp": 16, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.8421052631578947, + "f1": 0.9142857142857143 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 2, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 16, + 1, + 2 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 220, + "accuracy": 0.9454545454545454, + "macro_f1": 0.944095038434661, + "macro_f1_present_labels": 0.944095038434661, + "per_label": { + "AR": { + "support": 100, + "tp": 100, + "fp": 12, + "fn": 0, + "precision": 0.8928571428571429, + "recall": 1.0, + "f1": 0.9433962264150944 + }, + "DIFFUSION": { + "support": 60, + "tp": 48, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.8, + "f1": 0.8888888888888888 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 100, + 0, + 0 + ], + [ + 12, + 48, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "deu": { + "rows": 1, + "accuracy": 0.0, + "macro_f1": 0.0, + "macro_f1_present_labels": 0.0, + "per_label": { + "AR": { + "support": 1, + "tp": 0, + "fp": 0, + "fn": 1, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 0, + 0, + 1 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 256, + "accuracy": 0.96875, + "macro_f1": 0.9695893076174767, + "macro_f1_present_labels": 0.9695893076174767, + "per_label": { + "AR": { + "support": 112, + "tp": 107, + "fp": 3, + "fn": 5, + "precision": 0.9727272727272728, + "recall": 0.9553571428571429, + "f1": 0.963963963963964 + }, + "DIFFUSION": { + "support": 72, + "tp": 69, + "fp": 1, + "fn": 3, + "precision": 0.9857142857142858, + "recall": 0.9583333333333334, + "f1": 0.971830985915493 + }, + "BOTH": { + "support": 72, + "tp": 72, + "fp": 4, + "fn": 0, + "precision": 0.9473684210526315, + "recall": 1.0, + "f1": 0.972972972972973 + } + }, + "confusion": [ + [ + 107, + 1, + 4 + ], + [ + 3, + 69, + 0 + ], + [ + 0, + 0, + 72 + ] + ] + }, + "eng": { + "rows": 1, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 1, + "tp": 1, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 1, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 220, + "accuracy": 0.9772727272727273, + "macro_f1": 0.9743785567367925, + "macro_f1_present_labels": 0.9743785567367925, + "per_label": { + "AR": { + "support": 100, + "tp": 100, + "fp": 2, + "fn": 0, + "precision": 0.9803921568627451, + "recall": 1.0, + "f1": 0.9900990099009901 + }, + "DIFFUSION": { + "support": 60, + "tp": 58, + "fp": 3, + "fn": 2, + "precision": 0.9508196721311475, + "recall": 0.9666666666666667, + "f1": 0.9586776859504132 + }, + "BOTH": { + "support": 60, + "tp": 57, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.95, + "f1": 0.9743589743589743 + } + }, + "confusion": [ + [ + 100, + 0, + 0 + ], + [ + 2, + 58, + 0 + ], + [ + 0, + 3, + 57 + ] + ] + }, + "fr": { + "rows": 220, + "accuracy": 0.9272727272727272, + "macro_f1": 0.9236286001895273, + "macro_f1_present_labels": 0.9236286001895273, + "per_label": { + "AR": { + "support": 100, + "tp": 100, + "fp": 15, + "fn": 0, + "precision": 0.8695652173913043, + "recall": 1.0, + "f1": 0.9302325581395349 + }, + "DIFFUSION": { + "support": 60, + "tp": 45, + "fp": 1, + "fn": 15, + "precision": 0.9782608695652174, + "recall": 0.75, + "f1": 0.8490566037735849 + }, + "BOTH": { + "support": 60, + "tp": 59, + "fp": 0, + "fn": 1, + "precision": 1.0, + "recall": 0.9833333333333333, + "f1": 0.9915966386554622 + } + }, + "confusion": [ + [ + 100, + 0, + 0 + ], + [ + 15, + 45, + 0 + ], + [ + 0, + 1, + 59 + ] + ] + }, + "fra": { + "rows": 15, + "accuracy": 0.8666666666666667, + "macro_f1": 0.30952380952380953, + "macro_f1_present_labels": 0.9285714285714286, + "per_label": { + "AR": { + "support": 15, + "tp": 13, + "fp": 0, + "fn": 2, + "precision": 1.0, + "recall": 0.8666666666666667, + "f1": 0.9285714285714286 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 13, + 1, + 1 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "ja": { + "rows": 220, + "accuracy": 0.8636363636363636, + "macro_f1": 0.8454106280193235, + "macro_f1_present_labels": 0.8454106280193235, + "per_label": { + "AR": { + "support": 100, + "tp": 100, + "fp": 30, + "fn": 0, + "precision": 0.7692307692307693, + "recall": 1.0, + "f1": 0.8695652173913043 + }, + "DIFFUSION": { + "support": 60, + "tp": 30, + "fp": 0, + "fn": 30, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 60, + "tp": 60, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 100, + 0, + 0 + ], + [ + 30, + 30, + 0 + ], + [ + 0, + 0, + 60 + ] + ] + }, + "spa": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 256, + "accuracy": 0.91015625, + "macro_f1": 0.8999097602657681, + "macro_f1_present_labels": 0.8999097602657681, + "per_label": { + "AR": { + "support": 112, + "tp": 109, + "fp": 11, + "fn": 3, + "precision": 0.9083333333333333, + "recall": 0.9732142857142857, + "f1": 0.9396551724137931 + }, + "DIFFUSION": { + "support": 72, + "tp": 52, + "fp": 3, + "fn": 20, + "precision": 0.9454545454545454, + "recall": 0.7222222222222222, + "f1": 0.8188976377952756 + }, + "BOTH": { + "support": 72, + "tp": 72, + "fp": 9, + "fn": 0, + "precision": 0.8888888888888888, + "recall": 1.0, + "f1": 0.9411764705882353 + } + }, + "confusion": [ + [ + 109, + 3, + 0 + ], + [ + 11, + 52, + 9 + ], + [ + 0, + 0, + 72 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.7222222222222222, + "macro_f1": 0.7166666666666667, + "macro_f1_present_labels": 0.7166666666666667, + "per_label": { + "AR": { + "support": 6, + "tp": 4, + "fp": 0, + "fn": 2, + "precision": 1.0, + "recall": 0.6666666666666666, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 1, + "fn": 3, + "precision": 0.75, + "recall": 0.5, + "f1": 0.6 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 4, + "fn": 0, + "precision": 0.6, + "recall": 1.0, + "f1": 0.75 + } + }, + "confusion": [ + [ + 4, + 1, + 1 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.6111111111111112, + "macro_f1": 0.5888888888888889, + "macro_f1_present_labels": 0.5888888888888889, + "per_label": { + "AR": { + "support": 6, + "tp": 2, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.3333333333333333, + "f1": 0.5 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 1, + "fn": 3, + "precision": 0.75, + "recall": 0.5, + "f1": 0.6 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 6, + "fn": 0, + "precision": 0.5, + "recall": 1.0, + "f1": 0.6666666666666666 + } + }, + "confusion": [ + [ + 2, + 1, + 3 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 6, + 0, + 0 + ], + [ + 3, + 3, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + } + }, + "confusion": [ + [ + 6, + 0, + 0 + ], + [ + 0, + 3, + 3 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "short": { + "rows": 1339, + "accuracy": 0.9410007468259896, + "macro_f1": 0.938961528246535, + "macro_f1_present_labels": 0.938961528246535, + "per_label": { + "AR": { + "support": 619, + "tp": 614, + "fp": 70, + "fn": 5, + "precision": 0.8976608187134503, + "recall": 0.9919224555735057, + "f1": 0.9424405218726016 + }, + "DIFFUSION": { + "support": 360, + "tp": 290, + "fp": 7, + "fn": 70, + "precision": 0.9764309764309764, + "recall": 0.8055555555555556, + "f1": 0.882800608828006 + }, + "BOTH": { + "support": 360, + "tp": 356, + "fp": 2, + "fn": 4, + "precision": 0.994413407821229, + "recall": 0.9888888888888889, + "f1": 0.9916434540389972 + } + }, + "confusion": [ + [ + 614, + 3, + 2 + ], + [ + 70, + 290, + 0 + ], + [ + 0, + 4, + 356 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.7916666666666666, + "macro_f1": 0.7805555555555556, + "macro_f1_present_labels": 0.7805555555555556, + "per_label": { + "AR": { + "support": 8, + "tp": 7, + "fp": 1, + "fn": 1, + "precision": 0.875, + "recall": 0.875, + "f1": 0.875 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + } + }, + "confusion": [ + [ + 7, + 0, + 1 + ], + [ + 1, + 4, + 3 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.75, + "macro_f1": 0.7384615384615385, + "macro_f1_present_labels": 0.7384615384615385, + "per_label": { + "AR": { + "support": 8, + "tp": 6, + "fp": 1, + "fn": 2, + "precision": 0.8571428571428571, + "recall": 0.75, + "f1": 0.8 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 1, + "fn": 4, + "precision": 0.8, + "recall": 0.5, + "f1": 0.6153846153846154 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + } + }, + "confusion": [ + [ + 6, + 1, + 1 + ], + [ + 1, + 4, + 3 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.7083333333333334, + "macro_f1": 0.6971916971916973, + "macro_f1_present_labels": 0.6971916971916973, + "per_label": { + "AR": { + "support": 8, + "tp": 5, + "fp": 1, + "fn": 3, + "precision": 0.8333333333333334, + "recall": 0.625, + "f1": 0.7142857142857143 + }, + "DIFFUSION": { + "support": 8, + "tp": 4, + "fp": 1, + "fn": 4, + "precision": 0.8, + "recall": 0.5, + "f1": 0.6153846153846154 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 5, + "fn": 0, + "precision": 0.6153846153846154, + "recall": 1.0, + "f1": 0.7619047619047619 + } + }, + "confusion": [ + [ + 5, + 1, + 2 + ], + [ + 1, + 4, + 3 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "unspecified": { + "rows": 1339, + "accuracy": 0.9410007468259896, + "macro_f1": 0.938961528246535, + "macro_f1_present_labels": 0.938961528246535, + "per_label": { + "AR": { + "support": 619, + "tp": 614, + "fp": 70, + "fn": 5, + "precision": 0.8976608187134503, + "recall": 0.9919224555735057, + "f1": 0.9424405218726016 + }, + "DIFFUSION": { + "support": 360, + "tp": 290, + "fp": 7, + "fn": 70, + "precision": 0.9764309764309764, + "recall": 0.8055555555555556, + "f1": 0.882800608828006 + }, + "BOTH": { + "support": 360, + "tp": 356, + "fp": 2, + "fn": 4, + "precision": 0.994413407821229, + "recall": 0.9888888888888889, + "f1": 0.9916434540389972 + } + }, + "confusion": [ + [ + 614, + 3, + 2 + ], + [ + 70, + 290, + 0 + ], + [ + 0, + 4, + 356 + ] + ] + } + } + }, + "actual_tokens": { + "min": 10, + "max": 32768 + }, + "forward_batches_ms": [ + 6867.30867694132, + 9.979294030927122, + 10.217698058113456, + 9.577172924764454, + 8.845490054227412, + 9.165168972685933, + 9.64318704791367, + 9.60456405300647, + 8.76278302166611, + 8.968289010226727, + 9.47466993238777, + 9.677233058027923, + 9.375215042382479, + 9.823185973800719, + 9.662057971581817, + 9.179563028737903, + 11.730095953680575, + 9.293120936490595, + 8.677210076712072, + 9.391285013407469, + 9.813950979150832, + 9.575207019224763, + 9.551984025165439, + 8.885096991434693, + 9.53869300428778, + 8.742582984268665, + 8.80620803218335, + 8.974284981377423, + 8.370739058591425, + 8.279451983980834, + 8.752770954743028, + 8.849306963384151, + 8.838624926283956, + 8.66746692918241, + 8.844563039019704, + 8.776937960647047, + 8.734277915209532, + 8.659832063131034, + 8.778530056588352, + 8.63358296919614, + 8.674151031300426, + 8.512609987519681, + 9.786258917301893, + 8.91635799780488, + 8.802980999462306, + 8.703672094270587, + 8.956133970059454, + 9.44127095863223, + 9.805834968574345, + 8.607528056018054, + 8.948493050411344, + 8.621277054771781, + 8.823783951811492, + 8.64119699690491, + 8.930942974984646, + 8.808633079752326, + 8.573618018999696, + 8.275469997897744, + 8.803748991340399, + 8.867671946063638, + 8.73371004126966, + 9.083968005143106, + 9.118168032728136, + 8.670299896039069, + 10.530916973948479, + 9.835724020376801, + 9.535206016153097, + 10.862831026315689, + 9.53808892518282, + 9.13307594601065, + 8.736607967875898, + 8.652991964481771, + 8.785621030256152, + 8.867122000083327, + 8.924660040065646, + 8.9677240466699, + 8.791130967438221, + 9.663502918556333, + 10.761761921457946, + 10.900439927354455, + 13.626113068312407, + 13.141665025614202, + 26.179153006523848, + 10.871967067942023, + 10.509088053368032, + 10.136986966244876, + 10.065859998576343, + 10.550233069807291, + 9.864723077043891, + 10.230808984488249, + 10.376129997894168, + 10.429278947412968, + 10.588767006993294, + 10.126109002158046, + 95.11675895191729, + 94.12748797331005, + 93.1606040103361, + 94.69386609271169, + 95.08034691680223, + 95.58988595381379, + 90.89502296410501, + 92.33838389627635, + 94.54926906619221, + 90.11891495902091, + 91.88718197401613, + 91.07365796808153, + 92.71830692887306, + 90.55000694934279, + 90.10055009275675, + 91.15621098317206, + 89.35096103232354, + 92.38172706682235, + 297.9037309996784, + 295.3836730448529, + 301.1517010163516, + 310.11260591913015, + 285.07210698444396, + 281.12302406225353, + 293.4185180347413, + 301.85998196247965, + 317.9404450347647, + 308.69928805623204, + 284.06618908047676, + 290.1087219361216, + 289.306262973696, + 306.11998005770147, + 309.847544063814, + 316.40311097726226, + 296.2943450547755, + 291.6298599448055, + 1101.8928970443085, + 1187.7785919932649, + 1115.6065230024979, + 1103.6816139239818, + 1052.2410059347749, + 1127.9304079944268, + 1140.2252310654148, + 1127.1864819573238, + 1144.78641806636, + 1179.7228540526703, + 1068.4462479548529, + 1104.9326949287206, + 1157.1569900261238, + 1121.0174020379782, + 1086.3883190322667, + 1128.0008970061317, + 1118.4036789927632, + 1113.7385519687086, + 4351.286676013842, + 4134.929118910804, + 4212.978955009021, + 4117.138245957904, + 4257.8679820289835, + 4264.182258048095, + 4146.985196042806, + 4124.883470009081, + 4204.956557019614, + 4176.235282910056, + 4096.2329669855535, + 4119.872158044018, + 4040.9102799603716, + 4103.718529921025, + 4144.207939971238, + 4094.621406053193, + 4071.070461999625, + 4106.690975022502 + ], + "precision": { + "parameters": "float32", + "forward_autocast": null, + "device_type": "cuda" + }, + "data_files": [ + { + "file": "test.jsonl", + "sha256": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53" + }, + { + "file": "test.jsonl", + "sha256": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c" + } + ] +} diff --git a/evaluation/final-candidate-fp32.predictions.jsonl b/evaluation/final-candidate-fp32.predictions.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..0848d64a3d82608783f74f77d77c391847c9c2b3 --- /dev/null +++ b/evaluation/final-candidate-fp32.predictions.jsonl @@ -0,0 +1,1411 @@ +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-AR", "prediction": 0, "probabilities": [0.9998360872268677, 0.00016013211279641837, 3.6969790926377755e-06], "tokens": 39} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-DIFFUSION", "prediction": 1, "probabilities": [0.23572538793087006, 0.7504081130027771, 0.013866492547094822], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-BOTH", "prediction": 2, "probabilities": [1.2640257729401583e-09, 1.186483586934628e-07, 0.9999998807907104], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-AR", "prediction": 0, "probabilities": [0.9999780654907227, 2.1936077246209607e-05, 9.192669403468301e-10], "tokens": 38} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-DIFFUSION", "prediction": 0, "probabilities": [0.999169111251831, 0.0006901876768097281, 0.00014067151641938835], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-BOTH", "prediction": 2, "probabilities": [8.169430657289922e-06, 4.00315405824081e-11, 0.999991774559021], "tokens": 33} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-AR", "prediction": 0, "probabilities": [0.9999842643737793, 1.1923044439754449e-05, 3.868273324769689e-06], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-DIFFUSION", "prediction": 1, "probabilities": [0.00015512794198002666, 0.9998310804367065, 1.3728541489399504e-05], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-BOTH", "prediction": 2, "probabilities": [5.3100168706521345e-09, 0.00016401584434788674, 0.9998359680175781], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-AR", "prediction": 0, "probabilities": [0.9999849796295166, 8.934003972171922e-07, 1.4237007235351484e-05], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-DIFFUSION", "prediction": 1, "probabilities": [0.4381024241447449, 0.5618144869804382, 8.312499267049134e-05], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-BOTH", "prediction": 2, "probabilities": [2.488805961320395e-08, 8.331155640917132e-07, 0.9999991655349731], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-AR", "prediction": 0, "probabilities": [0.9999991655349731, 7.621541158187028e-07, 9.490911168086313e-08], "tokens": 38} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-DIFFUSION", "prediction": 0, "probabilities": [0.9998385906219482, 0.0001197460587718524, 4.164632991887629e-05], "tokens": 35} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-BOTH", "prediction": 2, "probabilities": [3.3139415656080473e-09, 1.438815502297075e-07, 0.9999998807907104], "tokens": 37} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-AR", "prediction": 0, "probabilities": [0.9999698400497437, 5.495074901773478e-07, 2.958358163596131e-05], "tokens": 39} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999858140945435, 1.1391032330720918e-06, 1.3040201338299084e-05], "tokens": 36} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-BOTH", "prediction": 2, "probabilities": [2.5044150087083494e-10, 3.5711373413960246e-08, 1.0], "tokens": 35} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-AR", "prediction": 0, "probabilities": [0.999951958656311, 1.2397757018334232e-05, 3.5599645343609154e-05], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-DIFFUSION", "prediction": 1, "probabilities": [3.7374698605083245e-13, 0.999998927116394, 1.0954656772810267e-06], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-BOTH", "prediction": 2, "probabilities": [9.948962487182911e-11, 9.448603011108503e-11, 1.0], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-AR", "prediction": 0, "probabilities": [0.9997926354408264, 0.00020534335635602474, 2.0246404801582685e-06], "tokens": 41} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-DIFFUSION", "prediction": 1, "probabilities": [4.571726027260969e-13, 1.0, 1.293686135994676e-08], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-BOTH", "prediction": 2, "probabilities": [4.5635762530027435e-11, 3.340978382837534e-09, 1.0], "tokens": 40} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-AR", "prediction": 0, "probabilities": [0.9999997615814209, 6.90324171159773e-08, 7.300502602447523e-08], "tokens": 46} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-DIFFUSION", "prediction": 1, "probabilities": [2.6471116854054344e-08, 0.9999967813491821, 3.2431794352305587e-06], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-BOTH", "prediction": 2, "probabilities": [2.624770789171754e-13, 1.571680147138821e-10, 1.0], "tokens": 43} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-AR", "prediction": 0, "probabilities": [1.0, 3.172638685100537e-08, 4.218901703684219e-12], "tokens": 47} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-DIFFUSION", "prediction": 1, "probabilities": [1.4491892061374578e-11, 1.0, 3.155674122012897e-08], "tokens": 39} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-BOTH", "prediction": 2, "probabilities": [2.8290824544141557e-12, 3.6612885878639645e-07, 0.9999996423721313], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-AR", "prediction": 0, "probabilities": [0.9999984502792358, 1.5759886764499242e-06, 6.485620374085954e-11], "tokens": 46} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-DIFFUSION", "prediction": 1, "probabilities": [7.789033307226134e-12, 0.9999998807907104, 1.3628239514673623e-07], "tokens": 40} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-BOTH", "prediction": 2, "probabilities": [7.608866314789964e-11, 4.777362880292735e-11, 1.0], "tokens": 44} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.4337262044250565e-08, 1.5426847710386937e-07], "tokens": 47} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-DIFFUSION", "prediction": 1, "probabilities": [1.6766002475243624e-10, 0.9999998807907104, 1.5419919918713276e-07], "tokens": 43} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-BOTH", "prediction": 2, "probabilities": [8.283121744057098e-09, 9.046101331477985e-05, 0.999909520149231], "tokens": 41} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-AR", "prediction": 0, "probabilities": [0.9999903440475464, 2.558955429776688e-07, 9.420661626791116e-06], "tokens": 76} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-DIFFUSION", "prediction": 0, "probabilities": [0.5562753677368164, 0.17553170025348663, 0.26819297671318054], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-BOTH", "prediction": 2, "probabilities": [8.19170722365925e-08, 2.0564383973464828e-08, 0.9999998807907104], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.8882367903170234e-07, 2.000811516467138e-09], "tokens": 75} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-DIFFUSION", "prediction": 0, "probabilities": [0.991119921207428, 0.008209466002881527, 0.0006706377025693655], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH", "prediction": 2, "probabilities": [0.21384473145008087, 3.3548049754861253e-11, 0.7861552834510803], "tokens": 70} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.2585287256783317e-09, 1.2860955678206665e-07], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-DIFFUSION", "prediction": 0, "probabilities": [0.9998570680618286, 1.1397149819458718e-06, 0.00014171232760418206], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-BOTH", "prediction": 2, "probabilities": [5.095773758512223e-06, 8.415808849804307e-08, 0.9999947547912598], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-AR", "prediction": 0, "probabilities": [1.0, 1.9909296433695545e-09, 4.1367940184500185e-09], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9999920129776001, 7.831812922631798e-07, 7.181823093560524e-06], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-BOTH", "prediction": 2, "probabilities": [0.0024163469206541777, 2.1814857973367907e-06, 0.9975814819335938], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-AR", "prediction": 0, "probabilities": [1.0, 1.3639687990174565e-10, 2.6031430344630735e-09], "tokens": 75} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-DIFFUSION", "prediction": 0, "probabilities": [0.9999890327453613, 8.16008238757604e-09, 1.0958440725516994e-05], "tokens": 72} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-BOTH", "prediction": 2, "probabilities": [1.3111568478052504e-05, 2.7220952603101978e-08, 0.9999868869781494], "tokens": 74} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-AR", "prediction": 0, "probabilities": [0.9999986886978149, 4.274882670074476e-09, 1.2904204140795628e-06], "tokens": 76} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999918937683105, 1.4864441766349046e-07, 7.990041012817528e-06], "tokens": 73} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-BOTH", "prediction": 2, "probabilities": [4.171065938862739e-06, 3.0299729392169183e-09, 0.9999958276748657], "tokens": 72} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-AR", "prediction": 0, "probabilities": [0.9999294281005859, 5.159766806173138e-05, 1.8926915799966082e-05], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-DIFFUSION", "prediction": 1, "probabilities": [1.20211045978591e-11, 0.9999984502792358, 1.564720719215984e-06], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-BOTH", "prediction": 2, "probabilities": [1.1345879791235802e-09, 8.784958871466131e-10, 1.0], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-AR", "prediction": 0, "probabilities": [0.999982476234436, 1.66905956575647e-05, 8.508449695909803e-07], "tokens": 30} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-DIFFUSION", "prediction": 1, "probabilities": [3.4919729868104765e-12, 1.0, 2.918543628993575e-08], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-BOTH", "prediction": 2, "probabilities": [1.0524486143692613e-10, 5.584339213271505e-09, 1.0], "tokens": 29} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-AR", "prediction": 0, "probabilities": [1.0, 9.528205779929522e-09, 5.270113589972425e-08], "tokens": 35} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-DIFFUSION", "prediction": 1, "probabilities": [6.206128944086231e-08, 0.9998502731323242, 0.00014965991431381553], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-BOTH", "prediction": 2, "probabilities": [1.4222460015256289e-12, 5.677204484300091e-09, 1.0], "tokens": 32} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-AR", "prediction": 0, "probabilities": [1.0, 5.393228263983474e-08, 2.7728405509186427e-12], "tokens": 36} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-DIFFUSION", "prediction": 1, "probabilities": [5.03627348158453e-12, 1.0, 5.111409606683992e-08], "tokens": 28} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-BOTH", "prediction": 2, "probabilities": [4.429904793684658e-12, 2.7696483329009425e-08, 1.0], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.3468322751796222e-07, 9.775492915142792e-11], "tokens": 35} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-DIFFUSION", "prediction": 1, "probabilities": [2.9299340731370194e-11, 1.0, 3.449224550422514e-08], "tokens": 29} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-BOTH", "prediction": 2, "probabilities": [2.280192701320516e-09, 2.7746195965794307e-11, 1.0], "tokens": 33} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-AR", "prediction": 0, "probabilities": [0.9999954700469971, 1.4665183512008184e-09, 4.557165539154084e-06], "tokens": 36} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-DIFFUSION", "prediction": 1, "probabilities": [2.1628758783975854e-09, 0.9999971389770508, 2.854693093468086e-06], "tokens": 32} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-BOTH", "prediction": 2, "probabilities": [5.329163954570504e-08, 2.835721772953548e-07, 0.9999997615814209], "tokens": 30} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-AR", "prediction": 0, "probabilities": [0.9995750784873962, 6.187248072819784e-05, 0.00036303125671111047], "tokens": 63} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-DIFFUSION", "prediction": 1, "probabilities": [0.0006595986778847873, 0.9934225678443909, 0.005917870439589024], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-BOTH", "prediction": 2, "probabilities": [2.942238008785125e-09, 8.004211622392177e-07, 0.9999991655349731], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-AR", "prediction": 0, "probabilities": [0.9999679327011108, 3.204222593922168e-05, 3.12537764557419e-08], "tokens": 62} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-DIFFUSION", "prediction": 0, "probabilities": [0.6688600778579712, 0.13585221767425537, 0.1952877640724182], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-BOTH", "prediction": 2, "probabilities": [6.960238351894077e-06, 1.2840932483992873e-10, 0.999993085861206], "tokens": 57} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-AR", "prediction": 0, "probabilities": [0.9997560381889343, 0.0002159973228117451, 2.8019212550134398e-05], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-DIFFUSION", "prediction": 1, "probabilities": [0.00038442417280748487, 0.9995917677879333, 2.380988371442072e-05], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-BOTH", "prediction": 2, "probabilities": [2.9468756324035894e-08, 0.005681000184267759, 0.9943190217018127], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-AR", "prediction": 0, "probabilities": [0.9999821186065674, 6.66468076815363e-06, 1.1158685083501041e-05], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9191965460777283, 0.08056202530860901, 0.00024142033362295479], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-BOTH", "prediction": 2, "probabilities": [1.209407827218456e-07, 0.00027418186073191464, 0.9997256398200989], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.3279617405714816e-08, 6.403798380461012e-08], "tokens": 62} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-DIFFUSION", "prediction": 1, "probabilities": [0.2990041673183441, 0.6338748931884766, 0.06712087243795395], "tokens": 59} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-BOTH", "prediction": 2, "probabilities": [9.319424343345872e-09, 7.373049356829142e-06, 0.9999926090240479], "tokens": 61} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-AR", "prediction": 0, "probabilities": [0.9997817873954773, 1.9010645928574377e-06, 0.00021628178365062922], "tokens": 63} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9955112934112549, 0.0001378718006890267, 0.004350831266492605], "tokens": 60} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-BOTH", "prediction": 2, "probabilities": [1.1689464674002892e-10, 7.913955073490797e-07, 0.9999991655349731], "tokens": 59} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-AR", "prediction": 0, "probabilities": [0.9999115467071533, 2.318461520189885e-05, 6.519466842291877e-05], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-DIFFUSION", "prediction": 1, "probabilities": [3.3071197690082155e-12, 0.9999473094940186, 5.264273568172939e-05], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-BOTH", "prediction": 2, "probabilities": [4.323862448640803e-12, 1.1280798517532276e-09, 1.0], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-AR", "prediction": 0, "probabilities": [0.9999903440475464, 6.405929980246583e-06, 3.255351884945412e-06], "tokens": 48} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-DIFFUSION", "prediction": 1, "probabilities": [1.8545925212221093e-11, 0.9999985694885254, 1.4761340025870595e-06], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-BOTH", "prediction": 2, "probabilities": [7.944814250926413e-11, 6.770622729135312e-09, 1.0], "tokens": 47} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-AR", "prediction": 0, "probabilities": [1.0, 1.394752757732931e-09, 1.5259434604786293e-08], "tokens": 53} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-DIFFUSION", "prediction": 1, "probabilities": [1.8333687901161966e-09, 0.9994704127311707, 0.000529622077010572], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-BOTH", "prediction": 2, "probabilities": [1.5085308219600635e-12, 6.233584315396001e-09, 1.0], "tokens": 50} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-AR", "prediction": 0, "probabilities": [1.0, 3.467602027740213e-08, 1.772917056219736e-11], "tokens": 54} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-DIFFUSION", "prediction": 1, "probabilities": [2.6586617022511527e-12, 1.0, 5.0630205805646256e-08], "tokens": 46} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-BOTH", "prediction": 2, "probabilities": [1.2493355959139474e-12, 9.277817980546388e-08, 0.9999998807907104], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.4418306193419994e-07, 1.9098643500026213e-10], "tokens": 53} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-DIFFUSION", "prediction": 1, "probabilities": [4.9720019767995893e-11, 0.9999959468841553, 4.01317311116145e-06], "tokens": 47} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-BOTH", "prediction": 2, "probabilities": [1.1344358785692066e-10, 2.3705796006745494e-11, 1.0], "tokens": 51} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-AR", "prediction": 0, "probabilities": [0.9999985694885254, 5.1125601530088716e-09, 1.4055337942409096e-06], "tokens": 54} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-DIFFUSION", "prediction": 1, "probabilities": [1.631258017553705e-09, 0.9999685287475586, 3.1479088647756726e-05], "tokens": 50} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-BOTH", "prediction": 2, "probabilities": [6.304544108992616e-10, 2.2066524252295494e-06, 0.9999977350234985], "tokens": 48} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-AR", "prediction": 0, "probabilities": [0.9979833364486694, 0.0018824780127033591, 0.00013423775089904666], "tokens": 38} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-DIFFUSION", "prediction": 1, "probabilities": [0.10638166964054108, 0.8626365065574646, 0.030981840565800667], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-BOTH", "prediction": 2, "probabilities": [3.915237856144671e-10, 2.217722112618503e-06, 0.9999977350234985], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-AR", "prediction": 0, "probabilities": [0.9821159839630127, 0.01788398064672947, 4.95170837666592e-08], "tokens": 37} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-DIFFUSION", "prediction": 1, "probabilities": [0.028336837887763977, 0.9313146471977234, 0.04034852236509323], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-BOTH", "prediction": 2, "probabilities": [1.9045366570935585e-05, 1.8909231958019745e-09, 0.9999809265136719], "tokens": 32} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-AR", "prediction": 0, "probabilities": [0.9958546161651611, 0.004045908804982901, 9.94268775684759e-05], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-DIFFUSION", "prediction": 1, "probabilities": [0.0003843271406367421, 0.9994943141937256, 0.00012132209667470306], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-BOTH", "prediction": 2, "probabilities": [7.247732969339893e-10, 0.001082785427570343, 0.9989172220230103], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-AR", "prediction": 0, "probabilities": [0.9996458292007446, 0.0001853485155152157, 0.00016891775885596871], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9634127020835876, 0.03296928480267525, 0.0036180694587528706], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-BOTH", "prediction": 2, "probabilities": [3.9922562478089674e-10, 2.6346961021772586e-05, 0.9999736547470093], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-AR", "prediction": 0, "probabilities": [0.9999909400939941, 8.95010271051433e-06, 1.62006230652878e-07], "tokens": 37} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-DIFFUSION", "prediction": 1, "probabilities": [0.004844906274229288, 0.9741979241371155, 0.020957129076123238], "tokens": 34} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-BOTH", "prediction": 2, "probabilities": [1.3836238821340885e-08, 5.79581392230466e-05, 0.9999420642852783], "tokens": 36} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-AR", "prediction": 0, "probabilities": [0.9998975992202759, 5.034882633481175e-05, 5.2116178267169744e-05], "tokens": 38} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9986996650695801, 8.820084622129798e-05, 0.0012121489271521568], "tokens": 35} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-BOTH", "prediction": 2, "probabilities": [2.9138365054137694e-09, 1.811283745212222e-08, 1.0], "tokens": 34} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-AR", "prediction": 0, "probabilities": [0.9971256852149963, 0.0002773120941128582, 0.0025969920679926872], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-DIFFUSION", "prediction": 1, "probabilities": [7.693798723118483e-12, 0.9993252754211426, 0.0006747188162989914], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-BOTH", "prediction": 2, "probabilities": [6.184006084986038e-11, 4.6402390263544646e-10, 1.0], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-AR", "prediction": 0, "probabilities": [0.9985779523849487, 0.0012981867184862494, 0.00012389422045089304], "tokens": 26} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-DIFFUSION", "prediction": 1, "probabilities": [9.067385731142963e-12, 0.9999991655349731, 7.851717214180098e-07], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-BOTH", "prediction": 2, "probabilities": [2.1650261444738916e-11, 5.787237356003061e-08, 1.0], "tokens": 25} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-AR", "prediction": 0, "probabilities": [1.0, 1.0878336453856718e-08, 3.307603790858593e-08], "tokens": 31} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-DIFFUSION", "prediction": 1, "probabilities": [4.4091027007198136e-07, 0.9385120272636414, 0.06148757413029671], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-BOTH", "prediction": 2, "probabilities": [4.450072580396147e-14, 6.191751555917335e-10, 1.0], "tokens": 28} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.90997888896527e-07, 2.2910447844415316e-11], "tokens": 32} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-DIFFUSION", "prediction": 1, "probabilities": [8.227219903608751e-13, 0.9999985694885254, 1.4335726064018672e-06], "tokens": 24} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-BOTH", "prediction": 2, "probabilities": [1.365563153006566e-12, 4.3270101390824323e-10, 1.0], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-AR", "prediction": 0, "probabilities": [0.9997501969337463, 0.0002498530375305563, 1.983171404873474e-09], "tokens": 31} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-DIFFUSION", "prediction": 1, "probabilities": [7.964470229160359e-12, 0.9999918937683105, 8.123369298118632e-06], "tokens": 25} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-BOTH", "prediction": 2, "probabilities": [1.8053701589249016e-10, 4.7210451693813216e-12, 1.0], "tokens": 29} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-AR", "prediction": 0, "probabilities": [0.9999961853027344, 1.0750220713617864e-09, 3.7929894460830837e-06], "tokens": 32} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-DIFFUSION", "prediction": 1, "probabilities": [2.480411653849046e-09, 0.9993659853935242, 0.0006340692634694278], "tokens": 28} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-BOTH", "prediction": 2, "probabilities": [1.9973651621540967e-09, 2.8732847212609158e-08, 1.0], "tokens": 26} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-AR", "prediction": 0, "probabilities": [0.9978556036949158, 0.0021406167652457952, 3.8081107049947605e-06], "tokens": 44} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-DIFFUSION", "prediction": 1, "probabilities": [0.015322122722864151, 0.9845234155654907, 0.00015441793948411942], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-BOTH", "prediction": 2, "probabilities": [1.537190641442976e-08, 3.97545809391886e-05, 0.9999603033065796], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-AR", "prediction": 0, "probabilities": [0.9973111152648926, 0.0026888803113251925, 3.8539001850912413e-11], "tokens": 43} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION", "prediction": 1, "probabilities": [0.3580162227153778, 0.641761839389801, 0.00022188712318893522], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-BOTH", "prediction": 2, "probabilities": [2.294384785272996e-06, 1.6072249087173418e-09, 0.9999977350234985], "tokens": 38} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-AR", "prediction": 0, "probabilities": [0.9981163740158081, 0.0018831348279491067, 4.991725290892646e-07], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-DIFFUSION", "prediction": 1, "probabilities": [0.013341043144464493, 0.9866580367088318, 8.746482080823625e-07], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-BOTH", "prediction": 1, "probabilities": [1.0188441734726439e-07, 0.7350020408630371, 0.26499783992767334], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-AR", "prediction": 0, "probabilities": [0.999943733215332, 5.593811874859966e-05, 3.85473242658918e-07], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9972105622291565, 0.0027892624493688345, 1.9729658617961832e-07], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-BOTH", "prediction": 2, "probabilities": [8.022389152984033e-08, 1.9629502276075073e-05, 0.9999802112579346], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-AR", "prediction": 0, "probabilities": [0.9999994039535522, 6.075925398363324e-07, 7.34933891521905e-09], "tokens": 43} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-DIFFUSION", "prediction": 0, "probabilities": [0.9515752792358398, 0.04841654375195503, 8.186712875613011e-06], "tokens": 40} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-BOTH", "prediction": 2, "probabilities": [1.0969893793344454e-07, 1.721911939966958e-05, 0.9999827146530151], "tokens": 42} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-AR", "prediction": 0, "probabilities": [0.9999988079071045, 4.747309674257849e-07, 7.165872375480831e-07], "tokens": 44} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999997615814209, 2.0840161596424878e-07, 4.593481151005108e-09], "tokens": 41} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-BOTH", "prediction": 2, "probabilities": [2.0039312431663348e-09, 5.678609227288689e-07, 0.9999994039535522], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-AR", "prediction": 0, "probabilities": [0.9999637603759766, 3.516750075505115e-05, 1.1299098332528956e-06], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-DIFFUSION", "prediction": 1, "probabilities": [8.958686114994563e-13, 0.9999994039535522, 6.209195362316677e-07], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-BOTH", "prediction": 2, "probabilities": [3.92324533970978e-10, 3.6934505454055966e-10, 1.0], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-AR", "prediction": 0, "probabilities": [0.9998866319656372, 0.00011305577936582267, 3.1071280659489275e-07], "tokens": 38} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-DIFFUSION", "prediction": 1, "probabilities": [8.46937797671643e-12, 1.0, 1.3469922954811864e-08], "tokens": 36} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-BOTH", "prediction": 2, "probabilities": [6.091706566263566e-12, 4.87868523357804e-10, 1.0], "tokens": 37} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-AR", "prediction": 0, "probabilities": [1.0, 3.0986229138108e-09, 3.0535753925420295e-09], "tokens": 42} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-DIFFUSION", "prediction": 1, "probabilities": [7.387397005231833e-08, 0.9999282360076904, 7.165941497078165e-05], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-BOTH", "prediction": 2, "probabilities": [2.7058078542509245e-13, 1.3254300434972777e-10, 1.0], "tokens": 39} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-AR", "prediction": 0, "probabilities": [1.0, 1.395256354896901e-08, 1.2019374211191813e-12], "tokens": 43} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-DIFFUSION", "prediction": 1, "probabilities": [2.309331371963541e-11, 1.0, 1.93481000110296e-08], "tokens": 35} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-BOTH", "prediction": 2, "probabilities": [8.882470280136001e-12, 4.722009094848545e-08, 1.0], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.3441525343059766e-07, 7.173922914049946e-12], "tokens": 42} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-DIFFUSION", "prediction": 1, "probabilities": [3.4927724774924673e-12, 1.0, 6.461744028740668e-09], "tokens": 36} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-BOTH", "prediction": 2, "probabilities": [4.507831052880107e-10, 5.664530042942539e-12, 1.0], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-AR", "prediction": 0, "probabilities": [0.999998927116394, 6.831085697101003e-10, 1.1196634659427218e-06], "tokens": 44} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-DIFFUSION", "prediction": 1, "probabilities": [7.049626993271829e-10, 0.9999997615814209, 1.7984848454943858e-07], "tokens": 40} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-BOTH", "prediction": 2, "probabilities": [8.187518005797756e-07, 1.5054160940053407e-07, 0.9999990463256836], "tokens": 38} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-AR", "prediction": 0, "probabilities": [0.9997934699058533, 1.4163028936309274e-05, 0.00019232854538131505], "tokens": 71} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-DIFFUSION", "prediction": 1, "probabilities": [1.9989425936728367e-07, 0.9991580247879028, 0.0008417886565439403], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-BOTH", "prediction": 2, "probabilities": [2.2286732181964197e-10, 8.68077313498361e-06, 0.9999912977218628], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-AR", "prediction": 0, "probabilities": [0.9995622038841248, 0.0004376003926154226, 2.1383523574058927e-07], "tokens": 70} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-DIFFUSION", "prediction": 1, "probabilities": [0.18378852307796478, 0.8125049471855164, 0.0037065644282847643], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-BOTH", "prediction": 2, "probabilities": [0.0005035846261307597, 3.3744063099305777e-10, 0.999496340751648], "tokens": 65} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-AR", "prediction": 0, "probabilities": [0.9997723698616028, 0.00021260482026264071, 1.5001010979176499e-05], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-DIFFUSION", "prediction": 1, "probabilities": [0.0002046627487288788, 0.9997637867927551, 3.1546023819828406e-05], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-BOTH", "prediction": 2, "probabilities": [1.4811162296624047e-09, 0.0024353619664907455, 0.9975646734237671], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-AR", "prediction": 0, "probabilities": [0.9999901056289673, 1.508877403466613e-06, 8.355392310477328e-06], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-DIFFUSION", "prediction": 1, "probabilities": [0.41557490825653076, 0.5748701691627502, 0.009554962627589703], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-BOTH", "prediction": 2, "probabilities": [8.510341764633722e-09, 0.001056263572536409, 0.9989437460899353], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 4.960643096296735e-08, 6.413112174641356e-08], "tokens": 70} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-DIFFUSION", "prediction": 1, "probabilities": [1.4569403901987243e-05, 0.9998486042022705, 0.00013684749137610197], "tokens": 67} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-BOTH", "prediction": 2, "probabilities": [1.5820086352036355e-09, 2.3127849999582395e-05, 0.9999768733978271], "tokens": 69} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-AR", "prediction": 0, "probabilities": [0.9999243021011353, 3.1135291465034243e-06, 7.264917803695425e-05], "tokens": 71} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999983310699463, 9.609521356424011e-08, 1.5330313090089476e-06], "tokens": 68} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-BOTH", "prediction": 2, "probabilities": [1.1246552578114688e-09, 1.0836425445859277e-07, 0.9999998807907104], "tokens": 67} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-AR", "prediction": 0, "probabilities": [0.9999940395355225, 2.8211381959408754e-06, 3.145118853353779e-06], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-DIFFUSION", "prediction": 1, "probabilities": [3.874943172322426e-13, 0.9999996423721313, 3.833548589682323e-07], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-BOTH", "prediction": 2, "probabilities": [1.3314212843340556e-08, 2.2270079114150576e-08, 1.0], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-AR", "prediction": 0, "probabilities": [0.999992847442627, 7.147263204387855e-06, 5.749962994627822e-08], "tokens": 77} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-DIFFUSION", "prediction": 1, "probabilities": [4.2736261578191215e-12, 0.9999998807907104, 8.669297812957666e-08], "tokens": 75} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-BOTH", "prediction": 2, "probabilities": [4.076210480263853e-09, 1.0386487503399877e-10, 1.0], "tokens": 76} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-AR", "prediction": 0, "probabilities": [1.0, 9.634069542130419e-09, 4.1518899429604517e-10], "tokens": 81} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-DIFFUSION", "prediction": 1, "probabilities": [6.7378533863404755e-09, 0.9999986886978149, 1.3173810202715686e-06], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-BOTH", "prediction": 2, "probabilities": [9.320786364952482e-10, 3.2612357259154123e-11, 1.0], "tokens": 78} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-AR", "prediction": 0, "probabilities": [1.0, 5.8325049678842333e-08, 9.818538126635246e-13], "tokens": 82} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-DIFFUSION", "prediction": 1, "probabilities": [9.140170564858607e-11, 1.0, 3.840035489588445e-08], "tokens": 74} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-BOTH", "prediction": 2, "probabilities": [3.912506152392581e-10, 1.7998129919760686e-07, 0.9999997615814209], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-AR", "prediction": 0, "probabilities": [1.0, 3.824451866307754e-08, 6.122975043654222e-11], "tokens": 81} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-DIFFUSION", "prediction": 1, "probabilities": [1.5383631868370884e-12, 1.0, 2.8393820628025424e-08], "tokens": 75} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-BOTH", "prediction": 2, "probabilities": [5.376576517335252e-09, 2.771643192733819e-10, 1.0], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-AR", "prediction": 0, "probabilities": [0.9999997615814209, 1.6057894597354405e-11, 2.5950521376216784e-07], "tokens": 83} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-DIFFUSION", "prediction": 1, "probabilities": [2.7493226445463037e-11, 1.0, 5.919180168234561e-08], "tokens": 79} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-BOTH", "prediction": 2, "probabilities": [1.1995183740509674e-05, 1.4983980918259476e-06, 0.9999864101409912], "tokens": 77} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-AR", "prediction": 0, "probabilities": [0.9870747327804565, 0.012568631209433079, 0.00035670105717144907], "tokens": 55} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-DIFFUSION", "prediction": 1, "probabilities": [0.0005624005571007729, 0.9992213249206543, 0.00021626900706905872], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-BOTH", "prediction": 2, "probabilities": [1.7506314176074511e-09, 8.818709829938598e-06, 0.9999911785125732], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-AR", "prediction": 0, "probabilities": [0.856113076210022, 0.14388692378997803, 1.7537997720751264e-09], "tokens": 54} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.006277744192630053, 0.9923650026321411, 0.0013572450261563063], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-BOTH", "prediction": 2, "probabilities": [0.00015850768249947578, 1.1676364408685913e-07, 0.9998413324356079], "tokens": 49} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-AR", "prediction": 0, "probabilities": [0.9998987913131714, 9.39395758905448e-05, 7.2446096055500675e-06], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-DIFFUSION", "prediction": 1, "probabilities": [1.458846213608922e-06, 0.9999958276748657, 2.6800219075084897e-06], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-BOTH", "prediction": 2, "probabilities": [7.341107721714479e-09, 0.00032956464565359056, 0.9996703863143921], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-AR", "prediction": 0, "probabilities": [0.9999825954437256, 1.705684007902164e-05, 3.324495025935903e-07], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-DIFFUSION", "prediction": 1, "probabilities": [1.7909940652316436e-05, 0.9999724626541138, 9.612930625735316e-06], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-BOTH", "prediction": 2, "probabilities": [1.5237110062571446e-07, 0.00018458202248439193, 0.9998152852058411], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-AR", "prediction": 0, "probabilities": [0.999990701675415, 9.145235708274413e-06, 1.5824939225694834e-07], "tokens": 54} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-DIFFUSION", "prediction": 1, "probabilities": [0.010563385672867298, 0.9859238862991333, 0.0035127068404108286], "tokens": 51} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-BOTH", "prediction": 2, "probabilities": [2.3675166271885928e-09, 4.652676943806e-05, 0.9999535083770752], "tokens": 53} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-AR", "prediction": 0, "probabilities": [0.9989977478981018, 0.0007150406599976122, 0.0002871883916668594], "tokens": 55} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9996345043182373, 0.0002688413660507649, 9.665458492236212e-05], "tokens": 52} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-BOTH", "prediction": 2, "probabilities": [6.184656342611561e-09, 2.587122708064271e-06, 0.9999973773956299], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-AR", "prediction": 0, "probabilities": [0.9998970031738281, 6.536900400533341e-06, 9.638978372095153e-05], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-DIFFUSION", "prediction": 1, "probabilities": [2.303621974165254e-13, 0.9999994039535522, 6.521090654132422e-07], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-BOTH", "prediction": 2, "probabilities": [3.252844382739539e-10, 4.994958266446581e-11, 1.0], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-AR", "prediction": 0, "probabilities": [0.9999951124191284, 4.698963493865449e-06, 1.831676001984306e-07], "tokens": 54} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-DIFFUSION", "prediction": 1, "probabilities": [4.976516594645819e-12, 0.9999998807907104, 1.411565051512298e-07], "tokens": 52} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-BOTH", "prediction": 2, "probabilities": [1.0433430785994346e-09, 1.9805547479823105e-10, 1.0], "tokens": 53} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-AR", "prediction": 0, "probabilities": [1.0, 1.136943250656941e-08, 1.5286197863062512e-09], "tokens": 58} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-DIFFUSION", "prediction": 1, "probabilities": [9.828807767675585e-10, 0.9999947547912598, 5.26311987414374e-06], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-BOTH", "prediction": 2, "probabilities": [1.9698307655163783e-12, 1.3409163779343203e-09, 1.0], "tokens": 55} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-AR", "prediction": 0, "probabilities": [0.9999988079071045, 1.1454603736638092e-06, 1.4121805938169252e-12], "tokens": 59} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-DIFFUSION", "prediction": 1, "probabilities": [1.5331155789333195e-10, 0.9999998807907104, 6.652625472725049e-08], "tokens": 51} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-BOTH", "prediction": 2, "probabilities": [1.4393548852797977e-11, 1.5940345932108357e-08, 1.0], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.6644499939720845e-07, 7.080144803772725e-11], "tokens": 58} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-DIFFUSION", "prediction": 1, "probabilities": [6.243459525420714e-13, 1.0, 4.706022949108046e-08], "tokens": 52} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-BOTH", "prediction": 2, "probabilities": [1.769499990977863e-09, 1.1593524751329909e-11, 1.0], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-AR", "prediction": 0, "probabilities": [0.9999994039535522, 2.898938034068266e-10, 5.418809791990498e-07], "tokens": 60} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-DIFFUSION", "prediction": 1, "probabilities": [9.730564964893773e-11, 0.9999995231628418, 4.186102273706638e-07], "tokens": 56} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-BOTH", "prediction": 2, "probabilities": [1.559962896635625e-07, 9.469652439975107e-08, 0.9999997615814209], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-AR", "prediction": 0, "probabilities": [0.9999920129776001, 3.067431180170388e-06, 4.844292107009096e-06], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-DIFFUSION", "prediction": 1, "probabilities": [0.03491135314106941, 0.961098849773407, 0.003989818040281534], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-BOTH", "prediction": 2, "probabilities": [6.623080972545381e-10, 2.074202711810358e-06, 0.9999979734420776], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-AR", "prediction": 0, "probabilities": [0.7690935730934143, 0.23090647161006927, 1.6219095400060723e-08], "tokens": 53} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-DIFFUSION", "prediction": 1, "probabilities": [0.029806669801473618, 0.9669182896614075, 0.0032750407699495554], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-BOTH", "prediction": 2, "probabilities": [0.00026127451565116644, 1.3671106469104188e-09, 0.9997387528419495], "tokens": 48} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-AR", "prediction": 0, "probabilities": [0.9997169375419617, 0.0002766449761111289, 6.396134267561138e-06], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-DIFFUSION", "prediction": 1, "probabilities": [4.526512566371821e-05, 0.9999388456344604, 1.5913301467662677e-05], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-BOTH", "prediction": 2, "probabilities": [5.07101503899321e-07, 0.0058647519908845425, 0.9941348433494568], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-AR", "prediction": 0, "probabilities": [0.9984022974967957, 0.0015820687403902411, 1.5556259313598275e-05], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-DIFFUSION", "prediction": 0, "probabilities": [0.5097509026527405, 0.489836186170578, 0.0004129379813093692], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-BOTH", "prediction": 2, "probabilities": [1.2302227503369068e-07, 0.000547451782040298, 0.9994524121284485], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-AR", "prediction": 0, "probabilities": [0.9999994039535522, 6.013655138303875e-07, 1.541843452912417e-08], "tokens": 53} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-DIFFUSION", "prediction": 0, "probabilities": [0.9982210993766785, 0.001229838584549725, 0.0005490032490342855], "tokens": 50} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-BOTH", "prediction": 2, "probabilities": [3.4364123990826556e-08, 9.739492816152051e-05, 0.999902606010437], "tokens": 52} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-AR", "prediction": 0, "probabilities": [0.9999654293060303, 1.504211741121253e-05, 1.9551005607354455e-05], "tokens": 54} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999983310699463, 3.002375592586759e-07, 1.2631941217478015e-06], "tokens": 51} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-BOTH", "prediction": 2, "probabilities": [1.9991639454985943e-09, 1.0155695662206199e-07, 0.9999998807907104], "tokens": 50} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-AR", "prediction": 0, "probabilities": [0.9999927282333374, 4.764691311720526e-06, 2.523586772440467e-06], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-DIFFUSION", "prediction": 1, "probabilities": [3.691256068166782e-12, 0.9999998807907104, 8.841549004046101e-08], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-BOTH", "prediction": 2, "probabilities": [2.143640376317535e-09, 1.8072950069836224e-07, 0.9999997615814209], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-AR", "prediction": 0, "probabilities": [0.9988462924957275, 0.0011504722060635686, 3.25826454172784e-06], "tokens": 53} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-DIFFUSION", "prediction": 1, "probabilities": [3.5318417027779825e-12, 1.0, 2.1474027889212266e-08], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-BOTH", "prediction": 2, "probabilities": [1.5962814403636116e-10, 2.820220057486722e-09, 1.0], "tokens": 52} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-AR", "prediction": 0, "probabilities": [0.9999970197677612, 2.925632998085348e-06, 1.5456322444151738e-08], "tokens": 58} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-DIFFUSION", "prediction": 1, "probabilities": [8.379280380665932e-09, 0.9999998807907104, 1.591938740830301e-07], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-BOTH", "prediction": 2, "probabilities": [8.693799803324853e-13, 9.897265229596997e-09, 1.0], "tokens": 55} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-AR", "prediction": 0, "probabilities": [0.9999980926513672, 1.8487229453967302e-06, 1.6013663919206556e-12], "tokens": 59} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-DIFFUSION", "prediction": 1, "probabilities": [2.882137887724734e-12, 1.0, 7.222785924909658e-09], "tokens": 51} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-BOTH", "prediction": 2, "probabilities": [2.0049429130808427e-11, 1.1852122042910196e-05, 0.9999881982803345], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-AR", "prediction": 0, "probabilities": [0.9999961853027344, 3.830953573924489e-06, 6.957292381093438e-12], "tokens": 58} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-DIFFUSION", "prediction": 1, "probabilities": [4.531870104879099e-12, 1.0, 1.1476671168964003e-08], "tokens": 52} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-BOTH", "prediction": 2, "probabilities": [3.056932040834681e-09, 2.2291900270143827e-10, 1.0], "tokens": 56} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-AR", "prediction": 0, "probabilities": [0.9999995231628418, 3.895745059878664e-09, 4.4811963562096935e-07], "tokens": 59} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-DIFFUSION", "prediction": 1, "probabilities": [7.457908422769943e-12, 1.0, 1.0241378411990354e-08], "tokens": 55} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-BOTH", "prediction": 2, "probabilities": [2.1697992451663595e-06, 0.0003806823515333235, 0.9996170997619629], "tokens": 53} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-AR", "prediction": 0, "probabilities": [0.9943076968193054, 0.005004978273063898, 0.0006873636739328504], "tokens": 42} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-DIFFUSION", "prediction": 1, "probabilities": [0.00034178001806139946, 0.9985743761062622, 0.0010838113958016038], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-BOTH", "prediction": 2, "probabilities": [6.267986663033298e-08, 1.2473676633817377e-06, 0.9999986886978149], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-AR", "prediction": 0, "probabilities": [0.890993058681488, 0.10900695621967316, 7.325571704797085e-09], "tokens": 42} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-DIFFUSION", "prediction": 1, "probabilities": [4.235095275362255e-06, 0.9999827146530151, 1.3049939298070967e-05], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-BOTH", "prediction": 2, "probabilities": [5.664180207531899e-06, 3.1959779267509703e-09, 0.9999942779541016], "tokens": 37} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-AR", "prediction": 0, "probabilities": [0.9338790774345398, 0.06588307023048401, 0.00023789283295627683], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-DIFFUSION", "prediction": 1, "probabilities": [8.723596920390264e-07, 0.9999966621398926, 2.4670005132065853e-06], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-BOTH", "prediction": 2, "probabilities": [2.453767411481067e-09, 0.007100205402821302, 0.9928997755050659], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-AR", "prediction": 0, "probabilities": [0.9990208148956299, 0.0009247713023796678, 5.4441414249595255e-05], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-DIFFUSION", "prediction": 1, "probabilities": [0.031456317752599716, 0.9683666825294495, 0.00017700667376630008], "tokens": 39} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-BOTH", "prediction": 2, "probabilities": [4.629471916928196e-09, 2.7185536964680068e-05, 0.9999728202819824], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-AR", "prediction": 0, "probabilities": [0.9999810457229614, 1.836570845625829e-05, 6.470730227192689e-07], "tokens": 41} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-DIFFUSION", "prediction": 1, "probabilities": [0.24650216102600098, 0.7455716133117676, 0.007926254533231258], "tokens": 38} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-BOTH", "prediction": 2, "probabilities": [3.690003635981043e-09, 5.330402927938849e-06, 0.9999946355819702], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-AR", "prediction": 0, "probabilities": [0.9997488856315613, 4.5683133066631854e-05, 0.00020547659369185567], "tokens": 43} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9991713762283325, 0.0008180067525245249, 1.0614614438964054e-05], "tokens": 40} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-BOTH", "prediction": 2, "probabilities": [1.2844818542134817e-09, 6.217957434273558e-06, 0.9999938011169434], "tokens": 39} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR", "prediction": 0, "probabilities": [0.9996397495269775, 0.00029394045122899115, 6.623043009312823e-05], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-DIFFUSION", "prediction": 1, "probabilities": [9.131427385067337e-12, 0.9994194507598877, 0.0005805380642414093], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-BOTH", "prediction": 2, "probabilities": [9.019716397384858e-12, 1.9537933937385787e-08, 1.0], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-AR", "prediction": 0, "probabilities": [0.996607780456543, 0.003371714847162366, 2.0496239812928252e-05], "tokens": 32} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-DIFFUSION", "prediction": 1, "probabilities": [7.549474934087641e-12, 0.9999940395355225, 6.0073948588978965e-06], "tokens": 30} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-BOTH", "prediction": 2, "probabilities": [8.57201950480585e-13, 1.2608377675249471e-09, 1.0], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-AR", "prediction": 0, "probabilities": [1.0, 2.794593889632324e-08, 4.0099028097984046e-08], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-DIFFUSION", "prediction": 1, "probabilities": [7.632776259924867e-07, 0.9410861134529114, 0.05891317501664162], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-BOTH", "prediction": 2, "probabilities": [3.7258421310215964e-14, 1.6275867764647955e-08, 1.0], "tokens": 35} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.747244138845417e-07, 3.4770734902433986e-11], "tokens": 39} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-DIFFUSION", "prediction": 1, "probabilities": [4.370530088011748e-13, 0.9999992847442627, 7.106585826477385e-07], "tokens": 31} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-BOTH", "prediction": 2, "probabilities": [1.2484655778806364e-13, 1.2431431439807739e-09, 1.0], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-AR", "prediction": 0, "probabilities": [0.9999985694885254, 1.4105263517194544e-06, 7.421805392482383e-10], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-DIFFUSION", "prediction": 1, "probabilities": [1.289143878452359e-11, 0.9999957084655762, 4.315041678637499e-06], "tokens": 32} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-BOTH", "prediction": 2, "probabilities": [5.1039682896769634e-11, 1.3616320571063767e-10, 1.0], "tokens": 36} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-AR", "prediction": 0, "probabilities": [0.9999983310699463, 2.4977730106456875e-08, 1.7082770682463888e-06], "tokens": 38} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-DIFFUSION", "prediction": 1, "probabilities": [3.10312192608464e-10, 0.999902606010437, 9.7435804491397e-05], "tokens": 34} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-BOTH", "prediction": 2, "probabilities": [7.655847067411514e-10, 2.5457008945295456e-08, 1.0], "tokens": 32} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-AR", "prediction": 0, "probabilities": [0.9999334812164307, 2.959520679723937e-05, 3.6920944694429636e-05], "tokens": 85} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-DIFFUSION", "prediction": 1, "probabilities": [1.5958154335748986e-06, 0.9999845027923584, 1.3939985365141183e-05], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-BOTH", "prediction": 2, "probabilities": [3.5116731744722074e-09, 0.013174794614315033, 0.986825168132782], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-AR", "prediction": 0, "probabilities": [0.9997054934501648, 0.00029442497179843485, 1.1766646546007564e-09], "tokens": 84} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-DIFFUSION", "prediction": 1, "probabilities": [0.1884293258190155, 0.8105537295341492, 0.0010169456945732236], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-BOTH", "prediction": 2, "probabilities": [0.0003889460931532085, 1.140633720808637e-08, 0.9996110796928406], "tokens": 79} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-AR", "prediction": 0, "probabilities": [0.9998921155929565, 0.00010587021824903786, 2.056824769169907e-06], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-DIFFUSION", "prediction": 1, "probabilities": [0.007232449483126402, 0.9927212595939636, 4.6223794925026596e-05], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-BOTH", "prediction": 2, "probabilities": [3.587265950955043e-07, 0.4971783459186554, 0.5028213262557983], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-AR", "prediction": 0, "probabilities": [0.9999758005142212, 2.2767306290916167e-05, 1.4100181715548388e-06], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9759190678596497, 0.02406160719692707, 1.929510290210601e-05], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-BOTH", "prediction": 1, "probabilities": [1.074280042701048e-08, 0.874710738658905, 0.12528929114341736], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 6.578444100568959e-08, 1.53306149996979e-08], "tokens": 84} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-DIFFUSION", "prediction": 1, "probabilities": [2.8306224066909635e-06, 0.9999818801879883, 1.5322146282414906e-05], "tokens": 81} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-BOTH", "prediction": 2, "probabilities": [3.110109858539545e-08, 0.0038744285702705383, 0.9961256384849548], "tokens": 83} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-AR", "prediction": 0, "probabilities": [0.9998859167098999, 6.778750048397342e-06, 0.0001072917366400361], "tokens": 85} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999829530715942, 3.6502822240436217e-06, 1.334843454969814e-05], "tokens": 82} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-BOTH", "prediction": 2, "probabilities": [2.9500451859121313e-08, 4.5658484850719105e-06, 0.9999954700469971], "tokens": 81} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-AR", "prediction": 0, "probabilities": [0.999131977558136, 1.656714994169306e-05, 0.0008514776127412915], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-DIFFUSION", "prediction": 1, "probabilities": [2.7992887955546886e-12, 0.9999105930328369, 8.940626867115498e-05], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-BOTH", "prediction": 2, "probabilities": [6.055146835326486e-12, 8.52467957335179e-11, 1.0], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-AR", "prediction": 0, "probabilities": [0.9985187649726868, 0.0014794187154620886, 1.7522206690046005e-06], "tokens": 39} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-DIFFUSION", "prediction": 1, "probabilities": [1.0683259832333647e-12, 0.9999998807907104, 1.2171828700502374e-07], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-BOTH", "prediction": 2, "probabilities": [7.764259070744406e-13, 6.477897329659754e-09, 1.0], "tokens": 38} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-AR", "prediction": 0, "probabilities": [0.9999997615814209, 1.46020067148811e-08, 2.8716476663248613e-07], "tokens": 44} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-DIFFUSION", "prediction": 1, "probabilities": [5.0018059027934214e-08, 0.9967237114906311, 0.003276255214586854], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-BOTH", "prediction": 2, "probabilities": [9.096225156565371e-14, 1.5597616254137137e-10, 1.0], "tokens": 41} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-AR", "prediction": 0, "probabilities": [1.0, 5.6916573676346616e-09, 1.1700494739752543e-10], "tokens": 45} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-DIFFUSION", "prediction": 1, "probabilities": [2.4435852646886858e-12, 0.9999997615814209, 2.4647820850987046e-07], "tokens": 37} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-BOTH", "prediction": 2, "probabilities": [2.638544222520467e-13, 8.180712107730415e-09, 1.0], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.1963356882915832e-07, 3.75311483280516e-11], "tokens": 44} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-DIFFUSION", "prediction": 1, "probabilities": [1.9929139935537243e-11, 0.9999977350234985, 2.312320475539309e-06], "tokens": 38} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-BOTH", "prediction": 2, "probabilities": [3.575916646125776e-11, 6.043793070176218e-12, 1.0], "tokens": 42} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-AR", "prediction": 0, "probabilities": [0.9999998807907104, 5.410307757358623e-09, 9.207970919078434e-08], "tokens": 45} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-DIFFUSION", "prediction": 1, "probabilities": [5.870257169249271e-09, 0.9999006986618042, 9.926321945386007e-05], "tokens": 41} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-BOTH", "prediction": 2, "probabilities": [4.830020272628133e-10, 7.701149939975949e-09, 1.0], "tokens": 39} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-AR", "prediction": 0, "probabilities": [0.9989832043647766, 0.00044196468661539257, 0.0005747796967625618], "tokens": 36} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-DIFFUSION", "prediction": 1, "probabilities": [0.019250065088272095, 0.8642169237136841, 0.11653300374746323], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-BOTH", "prediction": 2, "probabilities": [2.6810703657176305e-10, 1.1437384728196776e-07, 0.9999998807907104], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-AR", "prediction": 0, "probabilities": [0.9997751116752625, 0.00022484063811134547, 1.2024622408901564e-09], "tokens": 35} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-DIFFUSION", "prediction": 1, "probabilities": [0.24493476748466492, 0.7217556834220886, 0.03330961614847183], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-BOTH", "prediction": 2, "probabilities": [6.152930609459872e-08, 3.3216079192710524e-11, 0.9999998807907104], "tokens": 30} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-AR", "prediction": 0, "probabilities": [0.9999651908874512, 1.6997290003928356e-05, 1.7766265955287963e-05], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-DIFFUSION", "prediction": 1, "probabilities": [1.452639253329835e-06, 0.9999853372573853, 1.3257123100629542e-05], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-BOTH", "prediction": 2, "probabilities": [1.6348753462125387e-09, 0.000593324366491288, 0.9994066953659058], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-AR", "prediction": 0, "probabilities": [0.9999793767929077, 6.095958269725088e-06, 1.4533753528667148e-05], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-DIFFUSION", "prediction": 1, "probabilities": [0.0005079000839032233, 0.9993290901184082, 0.00016309498460032046], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-BOTH", "prediction": 2, "probabilities": [6.197539703656219e-10, 2.392364137904224e-07, 0.9999997615814209], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-AR", "prediction": 0, "probabilities": [0.9999996423721313, 2.8155858444733894e-07, 1.2702234641892574e-07], "tokens": 35} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-DIFFUSION", "prediction": 0, "probabilities": [0.6460634469985962, 0.332543283700943, 0.021393246948719025], "tokens": 32} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-BOTH", "prediction": 2, "probabilities": [6.239860295131905e-10, 1.9637768389202392e-07, 0.9999997615814209], "tokens": 34} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-AR", "prediction": 0, "probabilities": [0.9999198913574219, 3.5226287309342297e-06, 7.656202797079459e-05], "tokens": 36} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999383687973022, 5.166184564586729e-05, 1.0004060641222168e-05], "tokens": 33} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-BOTH", "prediction": 2, "probabilities": [1.5195249572497183e-11, 2.50911369903406e-07, 0.9999997615814209], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-AR", "prediction": 0, "probabilities": [0.9999722242355347, 2.765426688711159e-05, 1.5859153279507154e-07], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-DIFFUSION", "prediction": 1, "probabilities": [1.905081778094675e-12, 0.9999768733978271, 2.317496546311304e-05], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-BOTH", "prediction": 2, "probabilities": [1.4647813628698714e-08, 4.062785871616903e-11, 1.0], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-AR", "prediction": 0, "probabilities": [0.9999964237213135, 3.5983680390927475e-06, 1.601192956002251e-08], "tokens": 29} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-DIFFUSION", "prediction": 1, "probabilities": [5.998244002186626e-12, 0.9999995231628418, 4.235511426031735e-07], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-BOTH", "prediction": 2, "probabilities": [3.529607509045185e-11, 3.29799965115285e-10, 1.0], "tokens": 28} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-AR", "prediction": 0, "probabilities": [1.0, 1.0457045007683519e-08, 6.425080467664657e-10], "tokens": 34} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-DIFFUSION", "prediction": 1, "probabilities": [5.4314577937475406e-06, 0.9991949200630188, 0.0007996383355930448], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-BOTH", "prediction": 2, "probabilities": [7.505868214290057e-13, 2.6187041424208246e-10, 1.0], "tokens": 31} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 9.071838746876892e-08, 3.924286568444957e-13], "tokens": 35} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-DIFFUSION", "prediction": 1, "probabilities": [2.4598354464444583e-09, 0.999990701675415, 9.275521733798087e-06], "tokens": 27} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-BOTH", "prediction": 2, "probabilities": [3.477722233455327e-12, 4.800540853722168e-09, 1.0], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-AR", "prediction": 0, "probabilities": [0.9999995231628418, 4.5092244249644864e-07, 5.421298392843665e-12], "tokens": 34} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-DIFFUSION", "prediction": 1, "probabilities": [7.157439224636519e-11, 0.9999953508377075, 4.61375111626694e-06], "tokens": 28} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-BOTH", "prediction": 2, "probabilities": [1.7288145670590893e-07, 9.06465549323221e-13, 0.9999998807907104], "tokens": 32} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-AR", "prediction": 0, "probabilities": [0.9999966621398926, 1.20521348456748e-09, 3.387606284377398e-06], "tokens": 35} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-DIFFUSION", "prediction": 1, "probabilities": [5.238357303483099e-08, 0.9999566078186035, 4.339602674008347e-05], "tokens": 31} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-BOTH", "prediction": 2, "probabilities": [1.221760328462551e-07, 1.1112097908494434e-08, 0.9999998807907104], "tokens": 29} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-AR", "prediction": 0, "probabilities": [0.9997171759605408, 0.0002730230917222798, 9.742466318130028e-06], "tokens": 47} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-DIFFUSION", "prediction": 1, "probabilities": [0.002875235164538026, 0.9968928098678589, 0.0002318786719115451], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-BOTH", "prediction": 2, "probabilities": [1.6774498456939568e-09, 1.3429861382974195e-06, 0.9999986886978149], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-AR", "prediction": 0, "probabilities": [0.9293314814567566, 0.07066849619150162, 1.071288613374577e-09], "tokens": 46} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-DIFFUSION", "prediction": 0, "probabilities": [0.4977829158306122, 0.48255425691604614, 0.01966281421482563], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-BOTH", "prediction": 2, "probabilities": [2.3148768377723172e-05, 8.613331159423865e-10, 0.9999768733978271], "tokens": 41} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-AR", "prediction": 0, "probabilities": [0.999931812286377, 6.617129110964015e-05, 1.976544353965437e-06], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-DIFFUSION", "prediction": 1, "probabilities": [0.00010983520769514143, 0.9998741149902344, 1.5990432075341232e-05], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-BOTH", "prediction": 2, "probabilities": [3.216306510012146e-08, 0.04766792431473732, 0.9523321390151978], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-AR", "prediction": 0, "probabilities": [0.9999406337738037, 5.809197318740189e-05, 1.3421246194411651e-06], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-DIFFUSION", "prediction": 0, "probabilities": [0.6742967367172241, 0.32558542490005493, 0.00011787810944952071], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-BOTH", "prediction": 2, "probabilities": [1.7844847821635312e-09, 0.00034484872594475746, 0.9996551275253296], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-AR", "prediction": 0, "probabilities": [0.9999990463256836, 9.406733170180814e-07, 1.9658878969153193e-08], "tokens": 46} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-DIFFUSION", "prediction": 0, "probabilities": [0.9613969922065735, 0.031699568033218384, 0.006903345230966806], "tokens": 43} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-BOTH", "prediction": 2, "probabilities": [3.3982612279714886e-09, 5.692773356713587e-06, 0.9999942779541016], "tokens": 45} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-AR", "prediction": 0, "probabilities": [0.9999827146530151, 1.3303138075571042e-05, 3.993031441495987e-06], "tokens": 47} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999868869781494, 7.797814760124311e-06, 5.385438726079883e-06], "tokens": 44} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-BOTH", "prediction": 2, "probabilities": [2.2298105584184214e-09, 6.363593456626404e-07, 0.9999994039535522], "tokens": 43} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-AR", "prediction": 0, "probabilities": [0.9999938011169434, 9.146986030827975e-07, 5.230246188148158e-06], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-DIFFUSION", "prediction": 1, "probabilities": [1.6525310850626362e-12, 0.9999872446060181, 1.2732494724332355e-05], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-BOTH", "prediction": 2, "probabilities": [1.0713069320544832e-09, 7.362117165898852e-11, 1.0], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-AR", "prediction": 0, "probabilities": [0.9999966621398926, 3.021137899850146e-07, 2.9436216664180392e-06], "tokens": 54} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-DIFFUSION", "prediction": 1, "probabilities": [2.3922109085305898e-11, 0.9999949932098389, 4.9557866077520885e-06], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-BOTH", "prediction": 2, "probabilities": [1.2255454429066504e-08, 5.756244005755695e-13, 1.0], "tokens": 53} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-AR", "prediction": 0, "probabilities": [1.0, 1.3415201172151114e-09, 3.396221970319857e-08], "tokens": 59} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-DIFFUSION", "prediction": 1, "probabilities": [5.417263082563295e-07, 0.9992916584014893, 0.0007077977643348277], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-BOTH", "prediction": 2, "probabilities": [5.328130645021467e-11, 7.935507177585233e-13, 1.0], "tokens": 56} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-AR", "prediction": 0, "probabilities": [1.0, 4.00066468841942e-09, 6.804350052153074e-12], "tokens": 60} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-DIFFUSION", "prediction": 1, "probabilities": [2.107749139668158e-07, 0.9999462366104126, 5.3608313464792445e-05], "tokens": 52} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-BOTH", "prediction": 2, "probabilities": [4.592353840954244e-12, 6.330975743651379e-09, 1.0], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-AR", "prediction": 0, "probabilities": [1.0, 6.4785155018398655e-09, 3.2051025722346083e-10], "tokens": 59} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-DIFFUSION", "prediction": 1, "probabilities": [1.1982516541497734e-11, 0.9999990463256836, 9.707844128570287e-07], "tokens": 53} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-BOTH", "prediction": 2, "probabilities": [1.6462187168997389e-09, 9.87284798185939e-13, 1.0], "tokens": 57} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-AR", "prediction": 0, "probabilities": [0.9999998807907104, 7.894451758971854e-11, 1.3730806358580594e-07], "tokens": 60} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-DIFFUSION", "prediction": 1, "probabilities": [1.5453845980673009e-09, 0.999991774559021, 8.177240488294046e-06], "tokens": 56} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-BOTH", "prediction": 2, "probabilities": [2.991404016938759e-06, 1.5959129129328176e-08, 0.9999970197677612], "tokens": 54} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-AR", "prediction": 0, "probabilities": [0.9998557567596436, 0.0001285388570977375, 1.5722744137747213e-05], "tokens": 33} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-DIFFUSION", "prediction": 1, "probabilities": [0.011636962182819843, 0.9872480034828186, 0.0011150423670187593], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH", "prediction": 2, "probabilities": [3.77209996571537e-09, 2.0102582709569106e-07, 0.9999997615814209], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-AR", "prediction": 0, "probabilities": [0.9999709129333496, 2.9115382858435623e-05, 1.3287029254627214e-09], "tokens": 32} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-DIFFUSION", "prediction": 1, "probabilities": [0.01685762032866478, 0.980164110660553, 0.0029782408382743597], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-BOTH", "prediction": 2, "probabilities": [7.607949100929545e-06, 4.620711591130089e-10, 0.9999923706054688], "tokens": 27} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-AR", "prediction": 0, "probabilities": [0.9985224604606628, 0.0012709250440821052, 0.00020661331654991955], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-DIFFUSION", "prediction": 1, "probabilities": [5.580755532719195e-05, 0.9999434947967529, 6.737012654411956e-07], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-BOTH", "prediction": 2, "probabilities": [3.188366903827955e-09, 0.00010112714517163113, 0.9998989105224609], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-AR", "prediction": 0, "probabilities": [0.9999918937683105, 1.2435617691153311e-06, 6.95801372785354e-06], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-DIFFUSION", "prediction": 0, "probabilities": [0.6975463628768921, 0.3023037016391754, 0.00014989562623668462], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-BOTH", "prediction": 2, "probabilities": [2.5204542897228066e-09, 4.420600134835695e-07, 0.9999995231628418], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-AR", "prediction": 0, "probabilities": [0.9999983310699463, 1.4563811419066042e-06, 2.3952645733515965e-07], "tokens": 32} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-DIFFUSION", "prediction": 0, "probabilities": [0.9352694749832153, 0.06449869275093079, 0.0002317614998901263], "tokens": 29} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-BOTH", "prediction": 2, "probabilities": [8.175166321677807e-09, 2.974318931592279e-06, 0.9999970197677612], "tokens": 31} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-AR", "prediction": 0, "probabilities": [0.9999520778656006, 1.796405740606133e-05, 2.9882774470024742e-05], "tokens": 33} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999667406082153, 2.3315100406762213e-05, 9.899489668896422e-06], "tokens": 30} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-BOTH", "prediction": 2, "probabilities": [7.566953313942193e-11, 2.607960070122317e-08, 1.0], "tokens": 29} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-AR", "prediction": 0, "probabilities": [0.8649195432662964, 0.12685614824295044, 0.00822429545223713], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-DIFFUSION", "prediction": 1, "probabilities": [3.2308225105665e-12, 0.9999942779541016, 5.717932708648732e-06], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-BOTH", "prediction": 2, "probabilities": [8.660806793958642e-11, 2.269120669140534e-09, 1.0], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-AR", "prediction": 0, "probabilities": [0.9893820285797119, 0.01013887207955122, 0.0004791287938132882], "tokens": 29} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-DIFFUSION", "prediction": 1, "probabilities": [3.111113363457596e-12, 0.9999998807907104, 9.128231681643229e-08], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-BOTH", "prediction": 2, "probabilities": [1.5539948494680544e-10, 2.5616184728960434e-08, 1.0], "tokens": 28} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-AR", "prediction": 0, "probabilities": [1.0, 3.541386917049749e-08, 4.308501289074229e-08], "tokens": 34} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-DIFFUSION", "prediction": 1, "probabilities": [8.819775132451468e-09, 0.9999607801437378, 3.9158796425908804e-05], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-BOTH", "prediction": 2, "probabilities": [5.689832827983354e-13, 2.735235860029661e-08, 1.0], "tokens": 31} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-AR", "prediction": 0, "probabilities": [0.9999994039535522, 6.107686658651801e-07, 3.232860810650773e-12], "tokens": 35} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-DIFFUSION", "prediction": 1, "probabilities": [4.4617989858330986e-11, 0.9999998807907104, 7.202675789130808e-08], "tokens": 27} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-BOTH", "prediction": 2, "probabilities": [2.5218412427752135e-12, 3.046177354804058e-08, 1.0], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-AR", "prediction": 0, "probabilities": [0.9999918937683105, 8.134493327816017e-06, 3.890216621060816e-11], "tokens": 34} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-DIFFUSION", "prediction": 1, "probabilities": [5.0713122937140476e-12, 1.0, 1.0073892831030662e-08], "tokens": 28} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-BOTH", "prediction": 2, "probabilities": [8.222083014430837e-10, 2.9153290892480754e-11, 1.0], "tokens": 32} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-AR", "prediction": 0, "probabilities": [0.999893069267273, 1.1361866114611985e-08, 0.00010686105815693736], "tokens": 35} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-DIFFUSION", "prediction": 1, "probabilities": [1.3036091672802286e-07, 0.9999624490737915, 3.7391811929410324e-05], "tokens": 31} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-BOTH", "prediction": 2, "probabilities": [8.292123432340759e-09, 2.4836534606720306e-08, 1.0], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-AR", "prediction": 0, "probabilities": [0.9998176693916321, 0.00014553808432538062, 3.686199852381833e-05], "tokens": 30} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-DIFFUSION", "prediction": 0, "probabilities": [0.950019121170044, 0.046670544892549515, 0.0033103092573583126], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-BOTH", "prediction": 2, "probabilities": [8.709278298546508e-10, 3.361777771715424e-07, 0.9999996423721313], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-AR", "prediction": 0, "probabilities": [0.9984727501869202, 0.0015272872988134623, 5.147502424307504e-10], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-DIFFUSION", "prediction": 1, "probabilities": [0.001470372430048883, 0.9834113717079163, 0.015118224546313286], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-BOTH", "prediction": 2, "probabilities": [1.3145697721483884e-06, 9.80136305450685e-10, 0.9999986886978149], "tokens": 24} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-AR", "prediction": 0, "probabilities": [0.9677716493606567, 0.032030150294303894, 0.00019820829038508236], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-DIFFUSION", "prediction": 1, "probabilities": [4.4795161358024416e-08, 0.9999997615814209, 2.2459374804384424e-07], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-BOTH", "prediction": 2, "probabilities": [1.2072921828920613e-10, 0.0007701144786551595, 0.999229907989502], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-AR", "prediction": 0, "probabilities": [0.9981222748756409, 0.0016577397473156452, 0.00021995187853462994], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-DIFFUSION", "prediction": 1, "probabilities": [0.22820338606834412, 0.7717240452766418, 7.260704296641052e-05], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-BOTH", "prediction": 2, "probabilities": [1.8890673747495867e-10, 9.79525225375255e-07, 0.9999990463256836], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-AR", "prediction": 0, "probabilities": [0.9999980926513672, 1.909577122205519e-06, 2.976169177770771e-08], "tokens": 29} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-DIFFUSION", "prediction": 0, "probabilities": [0.9258145093917847, 0.07407711446285248, 0.00010835961438715458], "tokens": 26} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-BOTH", "prediction": 2, "probabilities": [1.1219170037435333e-09, 3.836566611425951e-05, 0.9999616146087646], "tokens": 28} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-AR", "prediction": 0, "probabilities": [0.9997497200965881, 0.0001870473352028057, 6.329567258944735e-05], "tokens": 30} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-DIFFUSION", "prediction": 0, "probabilities": [0.999339759349823, 0.000652673770673573, 7.491002634196775e-06], "tokens": 27} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-BOTH", "prediction": 2, "probabilities": [1.3572680479523314e-10, 8.111805982480291e-08, 0.9999998807907104], "tokens": 26} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-AR", "prediction": 0, "probabilities": [0.9999880790710449, 5.512749225999869e-07, 1.1376668226148468e-05], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-DIFFUSION", "prediction": 1, "probabilities": [4.955633885021793e-13, 0.9999991655349731, 7.901421668066178e-07], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-BOTH", "prediction": 2, "probabilities": [1.7698906229490774e-10, 1.3174510371527504e-09, 1.0], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-AR", "prediction": 0, "probabilities": [0.9998868703842163, 0.00011272454139543697, 4.1707562559167854e-07], "tokens": 56} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-DIFFUSION", "prediction": 1, "probabilities": [4.701507846233133e-12, 0.9999997615814209, 1.9375913495878194e-07], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-BOTH", "prediction": 2, "probabilities": [1.9701822118189938e-10, 2.994299599934358e-11, 1.0], "tokens": 55} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-AR", "prediction": 0, "probabilities": [1.0, 2.800739640207439e-09, 2.9766331621772224e-09], "tokens": 61} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-DIFFUSION", "prediction": 1, "probabilities": [2.353177930292105e-10, 0.9999864101409912, 1.3555132682085969e-05], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-BOTH", "prediction": 2, "probabilities": [2.5085621514070455e-12, 4.551096166593993e-10, 1.0], "tokens": 58} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 9.799833122769996e-08, 1.8049893177329857e-12], "tokens": 62} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-DIFFUSION", "prediction": 1, "probabilities": [3.609907373025578e-10, 0.9999996423721313, 3.7209073866506515e-07], "tokens": 54} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-BOTH", "prediction": 2, "probabilities": [7.635292571805952e-12, 2.210115468415097e-08, 1.0], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-AR", "prediction": 0, "probabilities": [1.0, 6.307809385930341e-09, 4.462724894488401e-12], "tokens": 61} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-DIFFUSION", "prediction": 1, "probabilities": [6.001862635357513e-12, 0.9999998807907104, 1.646066465355034e-07], "tokens": 55} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-BOTH", "prediction": 2, "probabilities": [7.505736587809508e-10, 1.0373138979724583e-11, 1.0], "tokens": 59} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-AR", "prediction": 0, "probabilities": [0.9999996423721313, 9.883895091267192e-11, 3.6726052599078685e-07], "tokens": 62} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-DIFFUSION", "prediction": 1, "probabilities": [3.0394534111799487e-10, 0.9999998807907104, 1.1818646328265459e-07], "tokens": 58} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-BOTH", "prediction": 2, "probabilities": [9.97401841118517e-08, 4.8296932675384596e-08, 0.9999998807907104], "tokens": 56} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-AR", "prediction": 0, "probabilities": [0.9999669790267944, 1.3284001397551037e-05, 1.9776029148488306e-05], "tokens": 28} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-DIFFUSION", "prediction": 1, "probabilities": [0.36719003319740295, 0.613067090511322, 0.019742852076888084], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-BOTH", "prediction": 2, "probabilities": [1.9538555218190368e-09, 1.219900980231614e-07, 0.9999998807907104], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-AR", "prediction": 0, "probabilities": [0.9871991276741028, 0.012800830416381359, 1.0457963384169489e-09], "tokens": 27} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-DIFFUSION", "prediction": 1, "probabilities": [0.03791165351867676, 0.86183100938797, 0.1002572774887085], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-BOTH", "prediction": 2, "probabilities": [1.3839068060406134e-06, 2.3393334491750295e-10, 0.9999985694885254], "tokens": 22} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-AR", "prediction": 0, "probabilities": [0.9999547004699707, 3.5146254958817735e-05, 1.011515450954903e-05], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-DIFFUSION", "prediction": 1, "probabilities": [2.2570395685761468e-07, 0.999990701675415, 9.07802223082399e-06], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-BOTH", "prediction": 2, "probabilities": [9.400284994853791e-10, 0.013071203604340553, 0.9869288206100464], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-AR", "prediction": 0, "probabilities": [0.9999139308929443, 2.7219475668971427e-05, 5.893575871596113e-05], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-DIFFUSION", "prediction": 1, "probabilities": [0.4017040431499481, 0.5864468812942505, 0.011849084869027138], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-BOTH", "prediction": 2, "probabilities": [2.5163200412237074e-09, 1.1378877388779074e-05, 0.9999886751174927], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-AR", "prediction": 0, "probabilities": [0.9999982118606567, 1.7492411643615924e-06, 5.653845391861978e-08], "tokens": 27} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-DIFFUSION", "prediction": 1, "probabilities": [0.16535170376300812, 0.6709348559379578, 0.16371344029903412], "tokens": 24} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-BOTH", "prediction": 2, "probabilities": [8.949744945141447e-09, 0.00019993928435724229, 0.9998000264167786], "tokens": 26} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-AR", "prediction": 0, "probabilities": [0.9996999502182007, 0.00027457173564471304, 2.550894168962259e-05], "tokens": 28} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9986674785614014, 0.0012751059839501977, 5.7434132031630725e-05], "tokens": 25} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-BOTH", "prediction": 2, "probabilities": [4.0855856892063613e-11, 3.036130991063146e-08, 1.0], "tokens": 24} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-AR", "prediction": 0, "probabilities": [0.9992522597312927, 2.022914486587979e-05, 0.0007274093804880977], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-DIFFUSION", "prediction": 1, "probabilities": [3.03080000332856e-12, 0.9995874762535095, 0.00041254438110627234], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-BOTH", "prediction": 2, "probabilities": [7.087590930821008e-11, 4.5777513724143404e-11, 1.0], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-AR", "prediction": 0, "probabilities": [0.9980781078338623, 0.0016642790287733078, 0.0002575959370005876], "tokens": 27} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-DIFFUSION", "prediction": 1, "probabilities": [2.7574875976349444e-12, 0.9999980926513672, 1.939491539815208e-06], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-BOTH", "prediction": 2, "probabilities": [1.3387830340855222e-12, 6.226276383358709e-09, 1.0], "tokens": 26} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-AR", "prediction": 0, "probabilities": [0.9999997615814209, 1.3283719013656992e-08, 2.2798987231453793e-07], "tokens": 32} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-DIFFUSION", "prediction": 1, "probabilities": [4.985953410141519e-07, 0.9710611701011658, 0.02893834374845028], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-BOTH", "prediction": 2, "probabilities": [1.1750886722006193e-13, 3.1283000495596625e-11, 1.0], "tokens": 29} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.5902127137887874e-07, 8.683100766182505e-11], "tokens": 33} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-DIFFUSION", "prediction": 1, "probabilities": [2.5205900734681652e-12, 0.9999939203262329, 6.132467206043657e-06], "tokens": 25} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-BOTH", "prediction": 2, "probabilities": [2.360749670835688e-13, 9.463324568415032e-10, 1.0], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-AR", "prediction": 0, "probabilities": [0.9994063377380371, 0.0005936925881542265, 9.42207534215811e-10], "tokens": 32} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-DIFFUSION", "prediction": 1, "probabilities": [2.6867135252683916e-11, 0.9999877214431763, 1.2279340808163397e-05], "tokens": 26} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-BOTH", "prediction": 2, "probabilities": [6.9050728676578466e-12, 8.533004164368307e-12, 1.0], "tokens": 30} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-AR", "prediction": 0, "probabilities": [0.9999706745147705, 2.2671076127522838e-09, 2.9327895390451886e-05], "tokens": 33} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-DIFFUSION", "prediction": 1, "probabilities": [3.3971478075045525e-08, 0.999648928642273, 0.0003509966190904379], "tokens": 29} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-BOTH", "prediction": 2, "probabilities": [4.611158954670458e-10, 5.823001103522074e-08, 1.0], "tokens": 27} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-AR", "prediction": 0, "probabilities": [0.9999659061431885, 2.3339529434451833e-05, 1.0698575351852924e-05], "tokens": 37} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-DIFFUSION", "prediction": 1, "probabilities": [0.00018656063184607774, 0.9991559982299805, 0.0006574366125278175], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-BOTH", "prediction": 2, "probabilities": [1.5941443720635107e-09, 1.1380441264918773e-06, 0.9999988079071045], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-AR", "prediction": 0, "probabilities": [0.9988080263137817, 0.0011919507523998618, 1.4062837561112929e-09], "tokens": 37} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-DIFFUSION", "prediction": 0, "probabilities": [0.9544429779052734, 0.04338381066918373, 0.002173194894567132], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-BOTH", "prediction": 2, "probabilities": [7.496317266486585e-06, 8.638156856477508e-11, 0.9999924898147583], "tokens": 32} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-AR", "prediction": 0, "probabilities": [0.9988402724266052, 0.0011420169612392783, 1.7718764866003767e-05], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-DIFFUSION", "prediction": 1, "probabilities": [0.00015760859241709113, 0.999819815158844, 2.265421426272951e-05], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-BOTH", "prediction": 2, "probabilities": [3.449902763463797e-08, 0.014614748768508434, 0.9853852391242981], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-AR", "prediction": 0, "probabilities": [0.9998443126678467, 0.00012556677393149585, 3.0134817279758863e-05], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-DIFFUSION", "prediction": 0, "probabilities": [0.987524151802063, 0.012414704076945782, 6.114922143751755e-05], "tokens": 34} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-BOTH", "prediction": 2, "probabilities": [3.0485506385957706e-07, 1.3015338481636718e-05, 0.9999866485595703], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-AR", "prediction": 0, "probabilities": [0.9999990463256836, 7.95257051322551e-07, 7.056463147137038e-08], "tokens": 36} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-DIFFUSION", "prediction": 0, "probabilities": [0.929898202419281, 0.0678045004606247, 0.0022972801234573126], "tokens": 33} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-BOTH", "prediction": 2, "probabilities": [8.1584801137069e-09, 1.3768966482530232e-06, 0.9999985694885254], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-AR", "prediction": 0, "probabilities": [0.9999793767929077, 2.697497336612287e-07, 2.0337831301731057e-05], "tokens": 38} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999969005584717, 2.762772965070326e-06, 3.508206987135054e-07], "tokens": 35} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-BOTH", "prediction": 2, "probabilities": [2.1050394760635527e-09, 1.9367452352980763e-07, 0.9999997615814209], "tokens": 34} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-AR", "prediction": 0, "probabilities": [0.9999352693557739, 8.838221219775733e-06, 5.593401874648407e-05], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-DIFFUSION", "prediction": 1, "probabilities": [8.284001956626597e-12, 0.9999829530715942, 1.7066575310309418e-05], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-BOTH", "prediction": 2, "probabilities": [2.6649751849738834e-10, 2.334999416042649e-10, 1.0], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-AR", "prediction": 0, "probabilities": [0.9995459914207458, 0.0004059600760228932, 4.807573714060709e-05], "tokens": 53} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-DIFFUSION", "prediction": 1, "probabilities": [1.1447618027127238e-11, 0.9999997615814209, 2.523081263916538e-07], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-BOTH", "prediction": 2, "probabilities": [6.663820051322489e-10, 1.9372932424932543e-10, 1.0], "tokens": 52} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-AR", "prediction": 0, "probabilities": [1.0, 4.994858748830211e-09, 1.2848874852977588e-08], "tokens": 58} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-DIFFUSION", "prediction": 1, "probabilities": [7.701692616990385e-09, 0.9998835325241089, 0.00011648800864350051], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-BOTH", "prediction": 2, "probabilities": [2.26400734097687e-12, 1.0577534681477374e-10, 1.0], "tokens": 55} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.1620733175732312e-07, 1.2934643807416268e-11], "tokens": 59} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-DIFFUSION", "prediction": 1, "probabilities": [1.3855483427249737e-10, 0.9999997615814209, 2.3166037976807274e-07], "tokens": 51} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-BOTH", "prediction": 2, "probabilities": [1.039162896357415e-12, 3.000781845230449e-08, 1.0], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.0250205068350624e-07, 1.4299265105321979e-09], "tokens": 58} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-DIFFUSION", "prediction": 1, "probabilities": [2.41199074602072e-12, 0.9999997615814209, 2.164167085538793e-07], "tokens": 52} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-BOTH", "prediction": 2, "probabilities": [1.8611942542712256e-10, 1.9547586205953538e-12, 1.0], "tokens": 56} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-AR", "prediction": 0, "probabilities": [0.9999985694885254, 1.5682022347363045e-09, 1.4179699974192772e-06], "tokens": 59} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-DIFFUSION", "prediction": 1, "probabilities": [6.957302095544904e-11, 0.9999990463256836, 9.460866863264528e-07], "tokens": 55} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-BOTH", "prediction": 2, "probabilities": [7.076333474742569e-08, 2.3116533043321397e-07, 0.9999996423721313], "tokens": 53} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-AR", "prediction": 0, "probabilities": [0.9997382760047913, 0.0002201462775701657, 4.1628649341873825e-05], "tokens": 30} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-DIFFUSION", "prediction": 1, "probabilities": [0.010711408220231533, 0.9841576814651489, 0.00513083953410387], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-BOTH", "prediction": 2, "probabilities": [4.414352261505883e-08, 2.4403629339531108e-08, 1.0], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-AR", "prediction": 0, "probabilities": [0.9983826875686646, 0.0016173351323232055, 9.83070291837862e-10], "tokens": 29} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-DIFFUSION", "prediction": 1, "probabilities": [0.03546256944537163, 0.8739738464355469, 0.09056362509727478], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-BOTH", "prediction": 2, "probabilities": [8.222755241149571e-07, 3.330614256613629e-11, 0.9999991655349731], "tokens": 24} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-AR", "prediction": 0, "probabilities": [0.9912484884262085, 0.00867383647710085, 7.767690112814307e-05], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-DIFFUSION", "prediction": 1, "probabilities": [0.0002186062338296324, 0.9996657371520996, 0.00011568002810236067], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-BOTH", "prediction": 2, "probabilities": [3.961208250302661e-09, 0.00032689046929590404, 0.9996731281280518], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-AR", "prediction": 0, "probabilities": [0.9999277591705322, 1.3905615560361184e-05, 5.82329485041555e-05], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-DIFFUSION", "prediction": 0, "probabilities": [0.8108327984809875, 0.1888270080089569, 0.00034015270648524165], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-BOTH", "prediction": 2, "probabilities": [3.608061627247139e-09, 1.1934085932807648e-06, 0.9999988079071045], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-AR", "prediction": 0, "probabilities": [0.9999914169311523, 8.44593159854412e-06, 1.2356329648355313e-07], "tokens": 29} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-DIFFUSION", "prediction": 1, "probabilities": [0.05776933953166008, 0.9418167471885681, 0.0004139360098633915], "tokens": 26} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-BOTH", "prediction": 2, "probabilities": [2.4094539696761785e-08, 1.7616195009395597e-06, 0.9999982118606567], "tokens": 28} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-AR", "prediction": 0, "probabilities": [0.9995099306106567, 0.0001467949041398242, 0.00034323552972637117], "tokens": 30} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999971389770508, 2.7013104499928886e-06, 7.217673925197232e-08], "tokens": 27} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-BOTH", "prediction": 2, "probabilities": [3.1219354879041816e-10, 4.4726444770049056e-08, 1.0], "tokens": 26} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-AR", "prediction": 0, "probabilities": [0.9999967813491821, 6.326471293505165e-07, 2.601436108307098e-06], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-DIFFUSION", "prediction": 1, "probabilities": [7.366741765213458e-12, 0.9999971389770508, 2.9032560178166023e-06], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-BOTH", "prediction": 2, "probabilities": [2.4971804624129845e-09, 1.7725410472735348e-08, 1.0], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-AR", "prediction": 0, "probabilities": [0.9999109506607056, 8.860122034093365e-05, 5.313186193234287e-07], "tokens": 55} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-DIFFUSION", "prediction": 1, "probabilities": [1.487675434230784e-11, 0.9999997615814209, 2.5307446094302577e-07], "tokens": 53} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-BOTH", "prediction": 2, "probabilities": [1.5784185070089052e-10, 2.3412006555645348e-08, 1.0], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-AR", "prediction": 0, "probabilities": [1.0, 6.2852794080470176e-09, 5.647752043813625e-09], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-DIFFUSION", "prediction": 1, "probabilities": [3.5664225350728884e-08, 0.9997547268867493, 0.0002452144108247012], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-BOTH", "prediction": 2, "probabilities": [4.077611512331991e-12, 1.610381445971143e-07, 0.9999998807907104], "tokens": 58} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 6.987760059473658e-08, 6.087988620173679e-12], "tokens": 62} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-DIFFUSION", "prediction": 1, "probabilities": [4.655783571172467e-12, 0.9999998807907104, 1.640878508624155e-07], "tokens": 54} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-BOTH", "prediction": 2, "probabilities": [3.381346418140918e-12, 6.930167728569359e-07, 0.9999992847442627], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-AR", "prediction": 0, "probabilities": [0.9999995231628418, 4.619614912826364e-07, 3.774168130799005e-11], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-DIFFUSION", "prediction": 1, "probabilities": [2.551859331484385e-12, 0.9999997615814209, 2.153385310066369e-07], "tokens": 55} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-BOTH", "prediction": 2, "probabilities": [1.2462429699322541e-10, 2.9208642171596466e-09, 1.0], "tokens": 59} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-AR", "prediction": 0, "probabilities": [0.9999997615814209, 5.255744728316358e-09, 2.1541185901696736e-07], "tokens": 61} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-DIFFUSION", "prediction": 1, "probabilities": [2.6251928408882463e-10, 0.9999990463256836, 9.763411981111858e-07], "tokens": 57} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-BOTH", "prediction": 2, "probabilities": [6.789696982423266e-08, 2.4257967652374646e-06, 0.9999974966049194], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-AR", "prediction": 0, "probabilities": [0.9999922513961792, 4.179300958639942e-06, 3.588599838622031e-06], "tokens": 59} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-DIFFUSION", "prediction": 1, "probabilities": [8.325695489475038e-06, 0.9999325275421143, 5.911484913667664e-05], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-BOTH", "prediction": 2, "probabilities": [1.0193573984906834e-08, 4.388740853755735e-05, 0.9999561309814453], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-AR", "prediction": 0, "probabilities": [0.7797635793685913, 0.22023580968379974, 5.587720011135389e-07], "tokens": 58} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0003463349712546915, 0.9995806813240051, 7.297322008525953e-05], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-BOTH", "prediction": 2, "probabilities": [0.0007940524956211448, 2.059506186924409e-06, 0.9992039799690247], "tokens": 53} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-AR", "prediction": 0, "probabilities": [0.9998753070831299, 7.03210971551016e-05, 5.436559149529785e-05], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-DIFFUSION", "prediction": 1, "probabilities": [3.562102790510835e-08, 0.9999998807907104, 6.880382841245591e-08], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-BOTH", "prediction": 2, "probabilities": [1.9743765733437613e-05, 0.08455637097358704, 0.9154238104820251], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-AR", "prediction": 0, "probabilities": [0.999984860420227, 4.59032798971748e-06, 1.0472950634721201e-05], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-DIFFUSION", "prediction": 1, "probabilities": [0.00013239191321190447, 0.9998643398284912, 3.286335413577035e-06], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-BOTH", "prediction": 2, "probabilities": [9.082013008310241e-08, 0.0014777745818719268, 0.9985221028327942], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-AR", "prediction": 0, "probabilities": [0.9999657869338989, 3.404136441531591e-05, 1.150526927062856e-07], "tokens": 58} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-DIFFUSION", "prediction": 1, "probabilities": [0.0019350590882822871, 0.997789740562439, 0.00027513279928825796], "tokens": 55} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-BOTH", "prediction": 2, "probabilities": [2.7588228590502695e-07, 4.1889234125847e-05, 0.9999579191207886], "tokens": 57} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-AR", "prediction": 0, "probabilities": [0.9999750852584839, 1.9865145077346824e-05, 5.064652668806957e-06], "tokens": 59} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9997870326042175, 0.00019592788885347545, 1.7070278772735037e-05], "tokens": 56} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-BOTH", "prediction": 2, "probabilities": [1.307748931367314e-08, 7.358298898907378e-05, 0.9999264478683472], "tokens": 55} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-AR", "prediction": 0, "probabilities": [0.9999688863754272, 2.7511365260579623e-05, 3.53797236130049e-06], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-DIFFUSION", "prediction": 1, "probabilities": [8.895892919871806e-13, 0.9999998807907104, 1.344314597417906e-07], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-BOTH", "prediction": 2, "probabilities": [4.271831777202806e-09, 1.4885181975898831e-08, 1.0], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-AR", "prediction": 0, "probabilities": [0.9999986886978149, 1.3051360383542487e-06, 2.1353429247028544e-08], "tokens": 62} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-DIFFUSION", "prediction": 1, "probabilities": [7.384925136688647e-12, 1.0, 1.7667488805273024e-08], "tokens": 60} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-BOTH", "prediction": 2, "probabilities": [2.5899311140165082e-08, 6.002164043561464e-11, 1.0], "tokens": 61} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.3106949836583226e-07, 1.913096703276551e-08], "tokens": 66} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-DIFFUSION", "prediction": 1, "probabilities": [1.3092334327780009e-08, 0.9999985694885254, 1.3651265362568665e-06], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-BOTH", "prediction": 2, "probabilities": [1.911583911995418e-11, 4.0420213842651265e-09, 1.0], "tokens": 63} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 9.552648094768301e-08, 1.2024417919698216e-12], "tokens": 67} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-DIFFUSION", "prediction": 1, "probabilities": [3.6164678890449053e-11, 1.0, 3.1927729349234824e-09], "tokens": 59} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-BOTH", "prediction": 2, "probabilities": [1.514481873243767e-10, 6.353018306981539e-07, 0.9999994039535522], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-AR", "prediction": 0, "probabilities": [1.0, 9.270137546479873e-09, 2.9982714228049545e-11], "tokens": 66} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-DIFFUSION", "prediction": 1, "probabilities": [2.495724330323079e-12, 1.0, 2.6659423557617856e-09], "tokens": 60} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-BOTH", "prediction": 2, "probabilities": [1.1459631465982056e-08, 2.4230407610903448e-11, 1.0], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-AR", "prediction": 0, "probabilities": [0.9999996423721313, 1.875268829110155e-10, 3.2389183957093337e-07], "tokens": 68} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-DIFFUSION", "prediction": 1, "probabilities": [4.50808862462182e-10, 0.9999998807907104, 1.0482870038686087e-07], "tokens": 64} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-BOTH", "prediction": 2, "probabilities": [3.2213463327934733e-06, 1.8333195157538285e-06, 0.9999949932098389], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-AR", "prediction": 0, "probabilities": [0.9994988441467285, 0.00048808983410708606, 1.3105182006256655e-05], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-DIFFUSION", "prediction": 1, "probabilities": [2.0931452127115335e-06, 0.9999814033508301, 1.6520576537004672e-05], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-BOTH", "prediction": 2, "probabilities": [5.571275440985346e-09, 0.00018487479246687144, 0.9998151659965515], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-AR", "prediction": 1, "probabilities": [0.016336295753717422, 0.9836636781692505, 5.962534466164016e-09], "tokens": 61} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0009429852943867445, 0.9988914132118225, 0.0001656789390835911], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-BOTH", "prediction": 2, "probabilities": [9.082889300771058e-05, 3.0227855774001e-08, 0.9999091625213623], "tokens": 56} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-AR", "prediction": 0, "probabilities": [0.9976174235343933, 0.002378583885729313, 3.887453203788027e-06], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-DIFFUSION", "prediction": 1, "probabilities": [3.6390049444889883e-06, 0.9999958276748657, 5.802352802675159e-07], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-BOTH", "prediction": 2, "probabilities": [1.9919451688110712e-07, 0.26849329471588135, 0.7315065264701843], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-AR", "prediction": 0, "probabilities": [0.9996411800384521, 0.00035728179500438273, 1.577178409206681e-06], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-DIFFUSION", "prediction": 1, "probabilities": [0.0012293721083551645, 0.9987683892250061, 2.2000569970259676e-06], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-BOTH", "prediction": 2, "probabilities": [1.0004935546703564e-08, 0.005755245219916105, 0.9942447543144226], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-AR", "prediction": 0, "probabilities": [0.9999982118606567, 1.7761558410711586e-06, 6.274248676163552e-09], "tokens": 61} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-DIFFUSION", "prediction": 1, "probabilities": [1.2233736015332397e-06, 0.9999971389770508, 1.6726319245208288e-06], "tokens": 58} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-BOTH", "prediction": 2, "probabilities": [7.074617336400024e-09, 0.0011344386730343103, 0.9988656044006348], "tokens": 60} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-AR", "prediction": 0, "probabilities": [0.9994708895683289, 0.0005142271402291954, 1.4907593140378594e-05], "tokens": 62} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9800655245780945, 0.019734367728233337, 0.00020012141612824053], "tokens": 59} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-BOTH", "prediction": 2, "probabilities": [2.5023394467638127e-09, 2.657003642525524e-05, 0.9999734163284302], "tokens": 58} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-AR", "prediction": 0, "probabilities": [0.9995561242103577, 0.0003452896198723465, 9.851288632489741e-05], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-DIFFUSION", "prediction": 1, "probabilities": [1.057164338708061e-11, 0.9997524619102478, 0.00024750595912337303], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-BOTH", "prediction": 2, "probabilities": [2.427973377905346e-10, 9.620916285868475e-10, 1.0], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-AR", "prediction": 0, "probabilities": [0.9998379945755005, 4.906497633783147e-05, 0.00011292712588328868], "tokens": 25} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-DIFFUSION", "prediction": 1, "probabilities": [1.5170429149002906e-11, 0.9999977350234985, 2.3210673134599347e-06], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-BOTH", "prediction": 2, "probabilities": [1.8587029484984363e-11, 9.131858463851117e-11, 1.0], "tokens": 24} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 5.6834443817876945e-09, 7.031533044710159e-08], "tokens": 30} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-DIFFUSION", "prediction": 1, "probabilities": [8.826137332107464e-07, 0.9559290409088135, 0.044070057570934296], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-BOTH", "prediction": 2, "probabilities": [1.721524263203697e-13, 5.572589611979595e-10, 1.0], "tokens": 27} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.7345125747378916e-07, 6.677577327890427e-12], "tokens": 31} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-DIFFUSION", "prediction": 1, "probabilities": [1.0043707185380946e-11, 0.999998927116394, 1.0142759947484592e-06], "tokens": 23} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-BOTH", "prediction": 2, "probabilities": [2.2468771634920337e-12, 1.3115341035430106e-09, 1.0], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-AR", "prediction": 0, "probabilities": [0.9999985694885254, 1.4179470326780574e-06, 3.123037245478244e-11], "tokens": 30} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-DIFFUSION", "prediction": 1, "probabilities": [1.4715842433510318e-10, 0.9999939203262329, 6.09644166615908e-06], "tokens": 24} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-BOTH", "prediction": 2, "probabilities": [1.2125213333380458e-10, 1.1062766154534831e-11, 1.0], "tokens": 28} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-AR", "prediction": 0, "probabilities": [0.9999971389770508, 1.9601360534693413e-09, 2.8217737053637393e-06], "tokens": 31} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-DIFFUSION", "prediction": 1, "probabilities": [2.3762568801544148e-08, 0.999525785446167, 0.0004742286982946098], "tokens": 27} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-BOTH", "prediction": 2, "probabilities": [4.473241688174312e-08, 2.5224622390851437e-08, 1.0], "tokens": 25} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-AR", "prediction": 0, "probabilities": [0.991786539554596, 0.003514423267915845, 0.0046991016715765], "tokens": 55} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-DIFFUSION", "prediction": 1, "probabilities": [0.1564403772354126, 0.7106246948242188, 0.13293497264385223], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-BOTH", "prediction": 2, "probabilities": [8.634618575698028e-10, 5.8866199026397226e-08, 1.0], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-AR", "prediction": 0, "probabilities": [0.9999442100524902, 5.577583215199411e-05, 5.230043509385496e-09], "tokens": 54} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0053754402324557304, 0.9230509996414185, 0.07157361507415771], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-BOTH", "prediction": 2, "probabilities": [4.2682572711782996e-06, 4.8424316639872345e-11, 0.9999957084655762], "tokens": 49} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-AR", "prediction": 0, "probabilities": [0.9999818801879883, 1.4152558833302464e-05, 3.89990509575e-06], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-DIFFUSION", "prediction": 1, "probabilities": [0.0004141198587603867, 0.999474823474884, 0.00011105988960480317], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-BOTH", "prediction": 2, "probabilities": [6.148890729917866e-09, 5.7481091062072664e-05, 0.9999425411224365], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-AR", "prediction": 0, "probabilities": [0.9999738931655884, 2.0907604266540147e-05, 5.229757789493306e-06], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9119816422462463, 0.08103504031896591, 0.006983282044529915], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-BOTH", "prediction": 2, "probabilities": [1.1836851498969736e-09, 2.933036057584104e-06, 0.9999970197677612], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-AR", "prediction": 0, "probabilities": [1.0, 2.1164463959166824e-08, 4.630112115933116e-08], "tokens": 54} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-DIFFUSION", "prediction": 0, "probabilities": [0.9595286250114441, 0.0035436039324849844, 0.036927685141563416], "tokens": 51} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-BOTH", "prediction": 2, "probabilities": [2.6202209291170675e-09, 1.3261366120786988e-06, 0.9999986886978149], "tokens": 53} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-AR", "prediction": 0, "probabilities": [0.9997950196266174, 5.8402893046149984e-05, 0.00014664742047898471], "tokens": 55} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9975529313087463, 1.6631573089398444e-05, 0.0024304608814418316], "tokens": 52} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-BOTH", "prediction": 2, "probabilities": [2.96992846982036e-10, 7.946065139208258e-09, 1.0], "tokens": 51} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-AR", "prediction": 0, "probabilities": [0.9999380111694336, 1.1395447472750675e-05, 5.050104300607927e-05], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-DIFFUSION", "prediction": 1, "probabilities": [1.2071706481653344e-12, 0.9999996423721313, 3.209268868431536e-07], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-BOTH", "prediction": 2, "probabilities": [4.059622416008324e-09, 1.8623165942699416e-06, 0.9999980926513672], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-AR", "prediction": 0, "probabilities": [0.9999605417251587, 3.933871630579233e-05, 9.543531120925763e-08], "tokens": 70} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-DIFFUSION", "prediction": 1, "probabilities": [2.369362518683804e-12, 1.0, 1.6118097079242943e-08], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-BOTH", "prediction": 2, "probabilities": [2.2974967761602016e-10, 6.263770058190232e-10, 1.0], "tokens": 69} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-AR", "prediction": 0, "probabilities": [1.0, 5.265612301741385e-08, 9.270239020864324e-10], "tokens": 75} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-DIFFUSION", "prediction": 1, "probabilities": [4.72201158174812e-08, 0.9999966621398926, 3.2420316529169213e-06], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-BOTH", "prediction": 2, "probabilities": [1.3055452552368507e-11, 2.742981120018584e-10, 1.0], "tokens": 72} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.152335636968928e-07, 1.0612526482600693e-11], "tokens": 76} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-DIFFUSION", "prediction": 1, "probabilities": [8.519673161400121e-11, 1.0, 2.3488938794002934e-08], "tokens": 68} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-BOTH", "prediction": 2, "probabilities": [3.41527431335642e-11, 3.688857361794362e-07, 0.9999996423721313], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-AR", "prediction": 0, "probabilities": [0.9999995231628418, 5.221855872150627e-07, 3.2628347246266287e-10], "tokens": 75} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-DIFFUSION", "prediction": 1, "probabilities": [2.040707446776535e-12, 1.0, 3.012377192135318e-08], "tokens": 69} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-BOTH", "prediction": 2, "probabilities": [6.681361242044659e-09, 1.1120462883873472e-10, 1.0], "tokens": 73} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-AR", "prediction": 0, "probabilities": [0.9999995231628418, 7.230718246376e-10, 4.271932141364232e-07], "tokens": 76} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-DIFFUSION", "prediction": 1, "probabilities": [1.6263740354460765e-11, 1.0, 4.719290203070159e-08], "tokens": 72} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-BOTH", "prediction": 2, "probabilities": [1.1576160829918081e-07, 2.8226588256075047e-05, 0.9999716281890869], "tokens": 70} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-AR", "prediction": 0, "probabilities": [0.9999107122421265, 8.774927846388891e-05, 1.5378021771539352e-06], "tokens": 34} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-DIFFUSION", "prediction": 0, "probabilities": [0.6129242777824402, 0.3090415596961975, 0.0780341774225235], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-BOTH", "prediction": 2, "probabilities": [2.724848116031353e-07, 7.971554083496812e-08, 0.9999996423721313], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-AR", "prediction": 0, "probabilities": [0.999601423740387, 0.0003985776857007295, 1.8078476216043526e-10], "tokens": 33} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-DIFFUSION", "prediction": 1, "probabilities": [0.005712681915611029, 0.9619194269180298, 0.03236783295869827], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-BOTH", "prediction": 2, "probabilities": [0.004448407329618931, 2.8610200875078817e-09, 0.9955515265464783], "tokens": 28} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-AR", "prediction": 0, "probabilities": [0.999618649482727, 0.00036240939516574144, 1.8981232642545365e-05], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-DIFFUSION", "prediction": 1, "probabilities": [1.608568595656834e-06, 0.9999802112579346, 1.812401387724094e-05], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-BOTH", "prediction": 2, "probabilities": [1.4320762886654848e-07, 0.00032505099079571664, 0.9996747970581055], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-AR", "prediction": 0, "probabilities": [0.9999871253967285, 5.046608748671133e-06, 7.824201020412147e-06], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9471916556358337, 0.05254748836159706, 0.0002607807982712984], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-BOTH", "prediction": 2, "probabilities": [7.190485007413372e-08, 2.021065057533633e-07, 0.9999996423721313], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.5489611605044047e-07, 4.422120625235948e-08], "tokens": 33} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-DIFFUSION", "prediction": 0, "probabilities": [0.9993566870689392, 0.00028490854310803115, 0.00035837446921505034], "tokens": 30} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-BOTH", "prediction": 2, "probabilities": [1.3774180843029171e-05, 5.188190243643476e-06, 0.9999809265136719], "tokens": 32} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-AR", "prediction": 0, "probabilities": [0.9999854564666748, 1.2438865269359667e-05, 2.136449438694399e-06], "tokens": 34} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999569654464722, 1.4198237295204308e-05, 2.888958806579467e-05], "tokens": 31} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-BOTH", "prediction": 2, "probabilities": [1.0190380761443407e-09, 9.012761092996868e-10, 1.0], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-AR", "prediction": 0, "probabilities": [0.9999337196350098, 4.029165575047955e-05, 2.5964454835047945e-05], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-DIFFUSION", "prediction": 1, "probabilities": [5.925248139360129e-12, 0.9988986253738403, 0.0011013366747647524], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-BOTH", "prediction": 2, "probabilities": [1.6587669822779039e-12, 2.256869829908581e-10, 1.0], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-AR", "prediction": 0, "probabilities": [0.9255744218826294, 0.07420364767313004, 0.00022184939007274806], "tokens": 32} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-DIFFUSION", "prediction": 1, "probabilities": [1.5066638258190435e-12, 0.9999994039535522, 6.176099986987538e-07], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-BOTH", "prediction": 2, "probabilities": [1.7782975612889995e-12, 2.4505020235210395e-09, 1.0], "tokens": 31} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 5.564862348705901e-08, 6.3799028282574e-08], "tokens": 37} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-DIFFUSION", "prediction": 1, "probabilities": [1.0744689404873498e-07, 0.9972184896469116, 0.0027814232744276524], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-BOTH", "prediction": 2, "probabilities": [8.573601084724963e-14, 4.260086380880601e-11, 1.0], "tokens": 34} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-AR", "prediction": 0, "probabilities": [1.0, 4.428460798067135e-08, 2.244950579599614e-11], "tokens": 38} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-DIFFUSION", "prediction": 1, "probabilities": [2.319189111588127e-12, 0.9999990463256836, 9.602799764252268e-07], "tokens": 30} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-BOTH", "prediction": 2, "probabilities": [1.073635212851809e-12, 1.1228442620137002e-09, 1.0], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-AR", "prediction": 0, "probabilities": [0.9974505305290222, 0.002549469005316496, 9.582195037438623e-10], "tokens": 37} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-DIFFUSION", "prediction": 1, "probabilities": [1.96356368664663e-11, 0.9999924898147583, 7.455002560163848e-06], "tokens": 31} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-BOTH", "prediction": 2, "probabilities": [3.9480432811878075e-12, 4.9667048251933466e-11, 1.0], "tokens": 35} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-AR", "prediction": 0, "probabilities": [0.9999494552612305, 2.470907745077966e-08, 5.0544844270916656e-05], "tokens": 38} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-DIFFUSION", "prediction": 1, "probabilities": [2.9138322310551246e-10, 0.999974250793457, 2.571268123574555e-05], "tokens": 34} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-BOTH", "prediction": 2, "probabilities": [6.474481367702012e-11, 8.975162835156425e-09, 1.0], "tokens": 32} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-AR", "prediction": 0, "probabilities": [0.9999344348907471, 5.855549534317106e-05, 7.001958692853805e-06], "tokens": 53} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-DIFFUSION", "prediction": 1, "probabilities": [2.820675217662938e-05, 0.9999171495437622, 5.468637027661316e-05], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-BOTH", "prediction": 2, "probabilities": [3.453536123743106e-09, 3.381065835128538e-05, 0.9999661445617676], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-AR", "prediction": 0, "probabilities": [0.9987014532089233, 0.0012985144276171923, 7.350401398653617e-10], "tokens": 52} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-DIFFUSION", "prediction": 0, "probabilities": [0.8681196570396423, 0.12766654789447784, 0.004213813692331314], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-BOTH", "prediction": 2, "probabilities": [2.007079501709086e-06, 3.0026792163795335e-10, 0.9999979734420776], "tokens": 47} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-AR", "prediction": 0, "probabilities": [0.9999842643737793, 1.536148374725599e-05, 3.766107283809106e-07], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-DIFFUSION", "prediction": 1, "probabilities": [1.0748686918304884e-06, 0.9999988079071045, 8.092093395362099e-08], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-BOTH", "prediction": 2, "probabilities": [1.937271676411001e-08, 0.06587632745504379, 0.934123694896698], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-AR", "prediction": 0, "probabilities": [0.9980890154838562, 0.0019057047320529819, 5.281229277898092e-06], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-DIFFUSION", "prediction": 1, "probabilities": [0.02991160750389099, 0.9700735211372375, 1.4842730706732254e-05], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-BOTH", "prediction": 2, "probabilities": [1.7670946927950126e-09, 0.004447847604751587, 0.995552122592926], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 7.246274691397048e-08, 9.73364233658458e-09], "tokens": 52} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-DIFFUSION", "prediction": 1, "probabilities": [0.06295933574438095, 0.9368131160736084, 0.00022748160699848086], "tokens": 49} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-BOTH", "prediction": 2, "probabilities": [4.50477488644907e-10, 7.16310678399168e-05, 0.99992835521698], "tokens": 51} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-AR", "prediction": 0, "probabilities": [0.9999775886535645, 7.012110927462345e-06, 1.5436276953550987e-05], "tokens": 53} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-DIFFUSION", "prediction": 0, "probabilities": [0.999995231628418, 8.280878773803124e-07, 3.910668510798132e-06], "tokens": 50} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-BOTH", "prediction": 2, "probabilities": [2.369521245881856e-10, 7.845421805541264e-07, 0.9999991655349731], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-AR", "prediction": 0, "probabilities": [0.9999934434890747, 6.381121693266323e-06, 1.734466081870778e-07], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-DIFFUSION", "prediction": 1, "probabilities": [2.537762209577077e-13, 0.9999995231628418, 5.016319164496963e-07], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-BOTH", "prediction": 2, "probabilities": [3.2105371694512996e-09, 7.595105794289125e-11, 1.0], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-AR", "prediction": 0, "probabilities": [0.9999840259552002, 1.5874276868999004e-05, 6.42495976421742e-08], "tokens": 43} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-DIFFUSION", "prediction": 1, "probabilities": [1.5149406252037978e-12, 1.0, 2.7120458767626587e-08], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-BOTH", "prediction": 2, "probabilities": [7.941602236938294e-11, 8.065345613594843e-10, 1.0], "tokens": 42} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-AR", "prediction": 0, "probabilities": [1.0, 4.548462939624187e-08, 2.9894406949892982e-09], "tokens": 48} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-DIFFUSION", "prediction": 1, "probabilities": [5.433106409213906e-09, 0.9999935626983643, 6.4761052271933295e-06], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-BOTH", "prediction": 2, "probabilities": [1.637554133092356e-12, 8.781766980270334e-12, 1.0], "tokens": 45} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-AR", "prediction": 0, "probabilities": [1.0, 3.512172597197605e-08, 8.031958887051716e-13], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-DIFFUSION", "prediction": 1, "probabilities": [4.191065203218436e-11, 1.0, 2.433307599858381e-08], "tokens": 41} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-BOTH", "prediction": 2, "probabilities": [1.5159356409055746e-11, 9.947705592594502e-08, 0.9999998807907104], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.0589987431994814e-07, 1.0492068151790601e-11], "tokens": 48} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-DIFFUSION", "prediction": 1, "probabilities": [4.831858437664982e-12, 1.0, 1.7790782180782116e-08], "tokens": 42} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-BOTH", "prediction": 2, "probabilities": [1.0888051349411398e-08, 3.3262728509064754e-12, 1.0], "tokens": 46} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-AR", "prediction": 0, "probabilities": [1.0, 1.7478442027041297e-09, 2.9727196704243397e-08], "tokens": 49} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-DIFFUSION", "prediction": 1, "probabilities": [1.328841814363102e-09, 0.9999990463256836, 9.215435170517594e-07], "tokens": 45} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-BOTH", "prediction": 2, "probabilities": [7.787537469994277e-06, 2.3309735297516454e-06, 0.9999898672103882], "tokens": 43} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-AR", "prediction": 1, "probabilities": [0.4534033536911011, 0.544163703918457, 0.0024329638108611107], "tokens": 30} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-DIFFUSION", "prediction": 1, "probabilities": [0.037269607186317444, 0.7666946649551392, 0.1960357129573822], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-BOTH", "prediction": 2, "probabilities": [1.3793349573631986e-08, 1.7847769129275548e-07, 0.9999998807907104], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-AR", "prediction": 0, "probabilities": [0.9999395608901978, 6.046493217581883e-05, 4.962269484209969e-10], "tokens": 29} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-DIFFUSION", "prediction": 0, "probabilities": [0.9255467653274536, 0.0668761357665062, 0.0075770108960568905], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-BOTH", "prediction": 2, "probabilities": [1.154796245828038e-05, 8.716475874415153e-10, 0.9999884366989136], "tokens": 24} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-AR", "prediction": 0, "probabilities": [0.999377429485321, 0.0005865001585334539, 3.61516467819456e-05], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-DIFFUSION", "prediction": 1, "probabilities": [0.0010531804291531444, 0.9986411929130554, 0.0003056191490031779], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-BOTH", "prediction": 2, "probabilities": [1.6718378070823547e-10, 0.021515408530831337, 0.978484570980072], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-AR", "prediction": 0, "probabilities": [0.9458780884742737, 0.05396350100636482, 0.00015842211723793298], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9993891716003418, 0.0003690984158311039, 0.00024168880190700293], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-BOTH", "prediction": 2, "probabilities": [1.940168470326853e-09, 0.01008954830467701, 0.9899104237556458], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-AR", "prediction": 0, "probabilities": [0.6982806921005249, 0.3017192482948303, 7.954914593710782e-08], "tokens": 29} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-DIFFUSION", "prediction": 0, "probabilities": [0.9846971035003662, 5.322239303495735e-05, 0.015249772928655148], "tokens": 26} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-BOTH", "prediction": 2, "probabilities": [6.009470254753069e-09, 0.000522022251971066, 0.9994780421257019], "tokens": 28} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-AR", "prediction": 0, "probabilities": [0.9999722242355347, 8.98885048172815e-07, 2.6858986529987305e-05], "tokens": 30} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999717473983765, 4.935687229590258e-06, 2.3334168872679584e-05], "tokens": 27} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-BOTH", "prediction": 2, "probabilities": [3.5452509816735756e-09, 1.8310954141842473e-10, 1.0], "tokens": 26} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-AR", "prediction": 0, "probabilities": [0.9999972581863403, 1.325244625149935e-06, 1.3827268503519008e-06], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-DIFFUSION", "prediction": 1, "probabilities": [1.9884146412740833e-12, 0.9999829530715942, 1.7016747733578086e-05], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-BOTH", "prediction": 2, "probabilities": [1.8790944356084083e-09, 2.1862944787898186e-10, 1.0], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-AR", "prediction": 0, "probabilities": [0.9999885559082031, 1.134119611378992e-05, 8.493758230088133e-08], "tokens": 26} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-DIFFUSION", "prediction": 1, "probabilities": [9.082326904441551e-12, 0.9999995231628418, 4.1969832409449737e-07], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-BOTH", "prediction": 2, "probabilities": [2.3855204270284425e-11, 1.4125185465729828e-09, 1.0], "tokens": 25} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-AR", "prediction": 0, "probabilities": [1.0, 2.6080482218304724e-09, 1.4802219450160692e-09], "tokens": 31} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-DIFFUSION", "prediction": 1, "probabilities": [4.628340377621498e-07, 0.9950646758079529, 0.004934860859066248], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-BOTH", "prediction": 2, "probabilities": [8.161552759577279e-14, 4.86244655650836e-10, 1.0], "tokens": 28} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-AR", "prediction": 0, "probabilities": [1.0, 4.1785082061096546e-08, 1.4700002012087848e-13], "tokens": 32} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-DIFFUSION", "prediction": 1, "probabilities": [1.3115418039805204e-12, 0.9999997615814209, 2.2167358793012681e-07], "tokens": 24} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-BOTH", "prediction": 2, "probabilities": [5.280161616098844e-13, 7.728253592631518e-09, 1.0], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-AR", "prediction": 0, "probabilities": [0.999998927116394, 1.0712410585256293e-06, 1.0486478005389266e-11], "tokens": 31} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-DIFFUSION", "prediction": 1, "probabilities": [1.0532107130867274e-10, 0.9999990463256836, 9.550732329444145e-07], "tokens": 25} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-BOTH", "prediction": 2, "probabilities": [3.1720985838035176e-09, 2.602727061651322e-12, 1.0], "tokens": 29} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-AR", "prediction": 0, "probabilities": [0.9999991655349731, 4.367819717021604e-10, 7.908010388746334e-07], "tokens": 32} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-DIFFUSION", "prediction": 1, "probabilities": [2.2882740324803308e-07, 0.9997097849845886, 0.0002899168466683477], "tokens": 28} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-BOTH", "prediction": 2, "probabilities": [7.999869922059588e-08, 9.92513236042214e-08, 0.9999997615814209], "tokens": 26} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-AR", "prediction": 0, "probabilities": [0.9999610185623169, 3.854600799968466e-05, 4.201298793304886e-07], "tokens": 58} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-DIFFUSION", "prediction": 1, "probabilities": [4.36809068560251e-06, 0.9999886751174927, 6.9162656473054085e-06], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-BOTH", "prediction": 2, "probabilities": [3.737300957595835e-08, 0.00010793524415930733, 0.999891996383667], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-AR", "prediction": 0, "probabilities": [0.9809277653694153, 0.01907220482826233, 2.5367863365488574e-09], "tokens": 57} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-DIFFUSION", "prediction": 1, "probabilities": [0.04104287549853325, 0.9586203694343567, 0.0003366924065630883], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-BOTH", "prediction": 2, "probabilities": [0.0012958013685420156, 7.093470255625789e-09, 0.998704195022583], "tokens": 52} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-AR", "prediction": 0, "probabilities": [0.9999551773071289, 4.412288035382517e-05, 6.703648978145793e-07], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-DIFFUSION", "prediction": 1, "probabilities": [0.0005926541052758694, 0.999405026435852, 2.3789900751580717e-06], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-BOTH", "prediction": 1, "probabilities": [1.9599539768933028e-08, 0.8544037938117981, 0.14559617638587952], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-AR", "prediction": 0, "probabilities": [0.9999593496322632, 4.051412906846963e-05, 1.3236426354978903e-07], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-DIFFUSION", "prediction": 1, "probabilities": [0.12783987820148468, 0.8721463680267334, 1.3743147064815275e-05], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-BOTH", "prediction": 2, "probabilities": [4.652580543051954e-08, 0.05256861820816994, 0.9474314451217651], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-AR", "prediction": 0, "probabilities": [1.0, 5.277496839539708e-08, 1.6037567718285572e-08], "tokens": 57} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-DIFFUSION", "prediction": 1, "probabilities": [0.3408498466014862, 0.6591221690177917, 2.8043372367392294e-05], "tokens": 54} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-BOTH", "prediction": 2, "probabilities": [3.093455447356064e-08, 0.0002560571883805096, 0.9997438788414001], "tokens": 56} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-AR", "prediction": 0, "probabilities": [0.9999967813491821, 7.710568752372637e-07, 2.5308638669230277e-06], "tokens": 58} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999698400497437, 1.508126842963975e-05, 1.4993688637332525e-05], "tokens": 55} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-BOTH", "prediction": 2, "probabilities": [1.5987779988790862e-08, 1.3760731235379353e-05, 0.9999862909317017], "tokens": 54} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-AR", "prediction": 0, "probabilities": [0.9999814033508301, 1.0785860467876773e-05, 7.846918379073031e-06], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-DIFFUSION", "prediction": 1, "probabilities": [1.5733901811629258e-12, 0.9999878406524658, 1.2140822036599275e-05], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-BOTH", "prediction": 2, "probabilities": [1.7627030501898844e-08, 3.392177164163179e-11, 1.0], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-AR", "prediction": 0, "probabilities": [0.9999996423721313, 1.806524352332417e-07, 1.1838136515507358e-07], "tokens": 46} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-DIFFUSION", "prediction": 1, "probabilities": [1.2063745835622086e-10, 0.9999988079071045, 1.1503274208735093e-06], "tokens": 44} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-BOTH", "prediction": 2, "probabilities": [4.2365494445917307e-10, 3.784439775444959e-12, 1.0], "tokens": 45} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 2.210035177085956e-08, 1.475197990430388e-07], "tokens": 50} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-DIFFUSION", "prediction": 1, "probabilities": [2.5107008241320727e-06, 0.9859625697135925, 0.014034898951649666], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-BOTH", "prediction": 2, "probabilities": [3.9212681712808006e-11, 2.0435207345737005e-11, 1.0], "tokens": 47} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-AR", "prediction": 0, "probabilities": [1.0, 3.71151891442878e-08, 1.6093711432962898e-11], "tokens": 51} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-DIFFUSION", "prediction": 1, "probabilities": [3.135098580742124e-08, 0.9999949932098389, 4.985594387107994e-06], "tokens": 43} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-BOTH", "prediction": 2, "probabilities": [1.6205226049947896e-10, 2.1961023277672354e-11, 1.0], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-AR", "prediction": 0, "probabilities": [1.0, 1.5696956623401093e-08, 3.6982794604512037e-10], "tokens": 50} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-DIFFUSION", "prediction": 1, "probabilities": [3.3896503659480715e-11, 0.9999988079071045, 1.2276918823772576e-06], "tokens": 44} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-BOTH", "prediction": 2, "probabilities": [5.270495506692896e-08, 1.1136009649137524e-12, 1.0], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-AR", "prediction": 0, "probabilities": [0.9999997615814209, 1.157079571045827e-10, 1.8462073114733357e-07], "tokens": 52} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-DIFFUSION", "prediction": 1, "probabilities": [1.2856266096150648e-07, 0.9997401833534241, 0.0002596671984065324], "tokens": 48} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-BOTH", "prediction": 2, "probabilities": [2.661301095940871e-06, 1.0970928165932037e-08, 0.9999973773956299], "tokens": 46} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-AR", "prediction": 0, "probabilities": [0.997786283493042, 0.0019971118308603764, 0.00021654512966051698], "tokens": 59} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION", "prediction": 1, "probabilities": [1.232252344607332e-07, 0.999985933303833, 1.3886413398722652e-05], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-BOTH", "prediction": 2, "probabilities": [4.079009074953177e-10, 1.3055943782092072e-05, 0.9999868869781494], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-AR", "prediction": 0, "probabilities": [0.9749583601951599, 0.025041650980710983, 6.177561573394996e-09], "tokens": 58} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-DIFFUSION", "prediction": 1, "probabilities": [1.4741458471689839e-05, 0.9999749660491943, 1.0247106729366351e-05], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-BOTH", "prediction": 2, "probabilities": [2.6922916731564328e-05, 4.44341026195616e-08, 0.9999730587005615], "tokens": 53} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-AR", "prediction": 0, "probabilities": [0.9970351457595825, 0.0029370435513556004, 2.777597364911344e-05], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-DIFFUSION", "prediction": 1, "probabilities": [5.4795658797957e-05, 0.9999386072158813, 6.54387167742243e-06], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-BOTH", "prediction": 2, "probabilities": [3.91763812501722e-08, 0.0072262100875377655, 0.9927737712860107], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-AR", "prediction": 0, "probabilities": [0.9928319454193115, 0.007117173634469509, 5.089138721814379e-05], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-DIFFUSION", "prediction": 1, "probabilities": [0.03768882900476456, 0.9622165560722351, 9.461375884711742e-05], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-BOTH", "prediction": 2, "probabilities": [2.3803363724539395e-09, 0.00036831371835432947, 0.9996316432952881], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-AR", "prediction": 0, "probabilities": [0.9999997615814209, 2.0834656311308208e-07, 3.056582897897897e-08], "tokens": 58} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-DIFFUSION", "prediction": 1, "probabilities": [0.00034838757710531354, 0.9996280670166016, 2.347172994632274e-05], "tokens": 55} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-BOTH", "prediction": 2, "probabilities": [2.4932673703403907e-09, 0.00048656162107363343, 0.9995133876800537], "tokens": 57} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-AR", "prediction": 0, "probabilities": [0.9999681711196899, 7.412766990455566e-06, 2.448985105729662e-05], "tokens": 59} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999945163726807, 1.7099364413297735e-06, 3.7792156035720836e-06], "tokens": 56} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-BOTH", "prediction": 2, "probabilities": [2.0533090783203534e-09, 3.872679315009009e-07, 0.9999996423721313], "tokens": 55} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-AR", "prediction": 0, "probabilities": [0.9999649524688721, 1.8609864127938636e-05, 1.6426640286226757e-05], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-DIFFUSION", "prediction": 1, "probabilities": [1.0164130821721518e-12, 0.9999972581863403, 2.710003855099785e-06], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-BOTH", "prediction": 2, "probabilities": [2.8361249113095788e-11, 1.981314140531154e-10, 1.0], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-AR", "prediction": 0, "probabilities": [0.9896185994148254, 0.010377855971455574, 3.549229404598009e-06], "tokens": 35} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-DIFFUSION", "prediction": 1, "probabilities": [1.761402005154289e-12, 0.9999998807907104, 1.6235475186476833e-07], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-BOTH", "prediction": 2, "probabilities": [6.0599650297810115e-12, 1.33376409916508e-09, 1.0], "tokens": 34} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-AR", "prediction": 0, "probabilities": [1.0, 7.0617711678266915e-09, 2.515649200063308e-08], "tokens": 40} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-DIFFUSION", "prediction": 1, "probabilities": [4.5135063686529975e-08, 0.9979137778282166, 0.0020862144883722067], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-BOTH", "prediction": 2, "probabilities": [1.2530627306172465e-13, 2.911351715262356e-10, 1.0], "tokens": 37} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-AR", "prediction": 0, "probabilities": [1.0, 4.4228389839418014e-08, 1.3934263898979227e-12], "tokens": 41} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-DIFFUSION", "prediction": 1, "probabilities": [3.3306211521394458e-12, 0.9999998807907104, 1.253140595736113e-07], "tokens": 33} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-BOTH", "prediction": 2, "probabilities": [1.2438859621141662e-12, 1.5740427627974896e-08, 1.0], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-AR", "prediction": 0, "probabilities": [1.0, 2.7944233593757417e-08, 6.7990606200663e-11], "tokens": 40} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-DIFFUSION", "prediction": 1, "probabilities": [4.42058629285258e-12, 0.9999998807907104, 1.4354179711517645e-07], "tokens": 34} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-BOTH", "prediction": 2, "probabilities": [5.382689294286536e-10, 5.6105944543083375e-12, 1.0], "tokens": 38} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-AR", "prediction": 0, "probabilities": [0.9999904632568359, 8.326193623453548e-10, 9.503263754595537e-06], "tokens": 41} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-DIFFUSION", "prediction": 1, "probabilities": [9.082789520498125e-10, 0.9999617338180542, 3.830531204584986e-05], "tokens": 37} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-BOTH", "prediction": 2, "probabilities": [7.4051627052540425e-09, 3.394589853655816e-08, 1.0], "tokens": 35} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-AR", "prediction": 0, "probabilities": [0.9983019828796387, 0.0015157056041061878, 0.00018233151058666408], "tokens": 31} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-DIFFUSION", "prediction": 1, "probabilities": [3.7175330362515524e-05, 0.9989726543426514, 0.0009901989251375198], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-BOTH", "prediction": 2, "probabilities": [1.7243670935584987e-09, 3.5360010031126876e-08, 1.0], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-AR", "prediction": 0, "probabilities": [0.9990653395652771, 0.000934657990001142, 2.728520465122841e-10], "tokens": 30} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0006099409656599164, 0.9971676468849182, 0.002222394337877631], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-BOTH", "prediction": 2, "probabilities": [5.867148956895107e-06, 7.345074548581465e-11, 0.999994158744812], "tokens": 25} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-AR", "prediction": 0, "probabilities": [0.9947925209999084, 0.005149231757968664, 5.8212568546878174e-05], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-DIFFUSION", "prediction": 1, "probabilities": [1.21826602139663e-07, 0.9999948740005493, 4.958897079632152e-06], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-BOTH", "prediction": 2, "probabilities": [1.2914846969636073e-09, 0.0008934906800277531, 0.9991064667701721], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-AR", "prediction": 0, "probabilities": [0.9999548196792603, 3.0738545319763944e-05, 1.4417422789847478e-05], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-DIFFUSION", "prediction": 1, "probabilities": [0.10991125553846359, 0.888731837272644, 0.001356945838779211], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-BOTH", "prediction": 2, "probabilities": [6.673495089870585e-09, 1.1945484175157617e-06, 0.9999988079071045], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-AR", "prediction": 0, "probabilities": [0.9999983310699463, 1.6064866485976381e-06, 1.283386978911949e-07], "tokens": 30} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-DIFFUSION", "prediction": 1, "probabilities": [0.09014122933149338, 0.7215444445610046, 0.1883142739534378], "tokens": 27} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-BOTH", "prediction": 2, "probabilities": [2.216133498933459e-09, 1.9239196547005122e-07, 0.9999997615814209], "tokens": 29} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-AR", "prediction": 0, "probabilities": [0.9999417066574097, 9.950609637598973e-06, 4.8447811423102394e-05], "tokens": 31} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-DIFFUSION", "prediction": 0, "probabilities": [0.999664306640625, 0.0002920826955232769, 4.3600182834779844e-05], "tokens": 28} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-BOTH", "prediction": 2, "probabilities": [3.309264667850087e-11, 1.3136627785570454e-08, 1.0], "tokens": 27} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-AR", "prediction": 0, "probabilities": [0.999985933303833, 7.453745183738647e-06, 6.604723239433952e-06], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-DIFFUSION", "prediction": 1, "probabilities": [8.139418501175444e-13, 0.9999960660934448, 3.9568562897329684e-06], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-BOTH", "prediction": 2, "probabilities": [1.534718185869366e-11, 1.9880333490540636e-10, 1.0], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-AR", "prediction": 0, "probabilities": [0.9999771118164062, 2.2183437977219e-05, 7.109533157745318e-07], "tokens": 42} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-DIFFUSION", "prediction": 1, "probabilities": [1.2263729528422251e-12, 0.9999998807907104, 1.6464652219383424e-07], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-BOTH", "prediction": 2, "probabilities": [4.534648004317354e-11, 8.938516701979815e-13, 1.0], "tokens": 41} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-AR", "prediction": 0, "probabilities": [1.0, 7.50276285543805e-09, 1.9862312683471828e-08], "tokens": 47} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-DIFFUSION", "prediction": 1, "probabilities": [3.236377921211897e-08, 0.9996084570884705, 0.00039154707337729633], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-BOTH", "prediction": 2, "probabilities": [1.5745583547936703e-13, 5.037109184619082e-12, 1.0], "tokens": 44} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-AR", "prediction": 0, "probabilities": [1.0, 4.6789306651362494e-08, 1.3779150346768243e-12], "tokens": 48} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-DIFFUSION", "prediction": 1, "probabilities": [2.6346644310670797e-10, 0.9999995231628418, 5.081935228190559e-07], "tokens": 40} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-BOTH", "prediction": 2, "probabilities": [1.2844197927464052e-11, 1.7351473502458248e-07, 0.9999998807907104], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-AR", "prediction": 0, "probabilities": [0.9999996423721313, 3.0062761879889877e-07, 2.2330190474484368e-10], "tokens": 47} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-DIFFUSION", "prediction": 1, "probabilities": [8.071780006543849e-13, 0.9999998807907104, 7.318334382944158e-08], "tokens": 41} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-BOTH", "prediction": 2, "probabilities": [9.249433441382848e-10, 2.5389846475265543e-12, 1.0], "tokens": 45} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-AR", "prediction": 0, "probabilities": [0.999998927116394, 2.7692886872543454e-10, 1.0318178738089046e-06], "tokens": 48} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-DIFFUSION", "prediction": 1, "probabilities": [8.628137315724871e-09, 0.9999734163284302, 2.6518191589275375e-05], "tokens": 44} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-BOTH", "prediction": 2, "probabilities": [5.610321522908635e-07, 4.9730132900549506e-08, 0.9999994039535522], "tokens": 42} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-AR", "prediction": 0, "probabilities": [0.9995373487472534, 0.0004308835486881435, 3.177411781507544e-05], "tokens": 62} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-DIFFUSION", "prediction": 1, "probabilities": [0.0010744512546807528, 0.9899210929870605, 0.00900451373308897], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-BOTH", "prediction": 2, "probabilities": [4.082590976395295e-08, 2.395734782112413e-06, 0.999997615814209], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-AR", "prediction": 0, "probabilities": [0.9995704293251038, 0.00042955143726430833, 5.216056475632058e-09], "tokens": 61} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-DIFFUSION", "prediction": 0, "probabilities": [0.9270869493484497, 0.07266737520694733, 0.0002456628135405481], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-BOTH", "prediction": 2, "probabilities": [0.00024401472182944417, 4.408158282842578e-09, 0.9997560381889343], "tokens": 56} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-AR", "prediction": 0, "probabilities": [0.9995896220207214, 0.00040804489981383085, 2.340920673304936e-06], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-DIFFUSION", "prediction": 1, "probabilities": [0.00225871242582798, 0.9977138042449951, 2.7432051865616813e-05], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-BOTH", "prediction": 2, "probabilities": [4.9609830909957964e-08, 0.0038494912441819906, 0.9961504936218262], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-AR", "prediction": 0, "probabilities": [0.9971010088920593, 0.002890952629968524, 7.954779903229792e-06], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-DIFFUSION", "prediction": 0, "probabilities": [0.726344883441925, 0.2735769748687744, 7.813259435351938e-05], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-BOTH", "prediction": 2, "probabilities": [6.088925808001022e-09, 4.134389200771693e-06, 0.9999958276748657], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.2333570964528917e-07, 2.003486088142381e-08], "tokens": 61} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-DIFFUSION", "prediction": 1, "probabilities": [0.06668440252542496, 0.9331163763999939, 0.00019920759950764477], "tokens": 58} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-BOTH", "prediction": 2, "probabilities": [2.2205839389499715e-09, 2.3666427750868024e-06, 0.999997615814209], "tokens": 60} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-AR", "prediction": 0, "probabilities": [0.9999672174453735, 4.0202439777203836e-06, 2.8761724024661817e-05], "tokens": 62} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999915361404419, 5.448317097034305e-06, 3.0052356123633217e-06], "tokens": 59} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-BOTH", "prediction": 2, "probabilities": [4.577726198107257e-09, 1.0398339327366557e-05, 0.9999896287918091], "tokens": 58} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-AR", "prediction": 0, "probabilities": [0.9996273517608643, 0.000371179252397269, 1.4229642602003878e-06], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-DIFFUSION", "prediction": 1, "probabilities": [3.916314429870671e-13, 0.9999995231628418, 4.5239082169246103e-07], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-BOTH", "prediction": 2, "probabilities": [2.8860849821121803e-10, 2.32682917378213e-09, 1.0], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR", "prediction": 0, "probabilities": [0.9998751878738403, 0.00012465703184716403, 1.7623361259211379e-07], "tokens": 43} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-DIFFUSION", "prediction": 1, "probabilities": [6.851202547994428e-13, 1.0, 2.598248372009948e-08], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-BOTH", "prediction": 2, "probabilities": [3.5134135184522464e-11, 5.410627945678925e-09, 1.0], "tokens": 42} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.34076415747586e-07, 4.925925001231235e-09], "tokens": 48} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-DIFFUSION", "prediction": 1, "probabilities": [1.880889666239227e-09, 0.9999973773956299, 2.6192744826403214e-06], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-BOTH", "prediction": 2, "probabilities": [5.310259068407042e-13, 8.966855036263155e-11, 1.0], "tokens": 45} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.7789143669233454e-07, 2.2079945294004033e-13], "tokens": 49} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-DIFFUSION", "prediction": 1, "probabilities": [8.543068162614187e-12, 1.0, 3.2428186358401945e-08], "tokens": 41} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-BOTH", "prediction": 2, "probabilities": [3.0737055695811977e-12, 5.295470373312128e-07, 0.9999995231628418], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-AR", "prediction": 0, "probabilities": [0.9998238682746887, 0.0001761512248776853, 3.596825268181725e-11], "tokens": 48} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-DIFFUSION", "prediction": 1, "probabilities": [1.4879076703361305e-12, 1.0, 2.4085025529529958e-08], "tokens": 42} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-BOTH", "prediction": 2, "probabilities": [6.315559186731434e-10, 1.8347626196124622e-10, 1.0], "tokens": 46} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-AR", "prediction": 0, "probabilities": [0.9999998807907104, 2.467348103607492e-09, 1.0913025505487894e-07], "tokens": 49} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-DIFFUSION", "prediction": 1, "probabilities": [3.115735100478467e-11, 0.9999998807907104, 1.2690735218257032e-07], "tokens": 45} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-BOTH", "prediction": 2, "probabilities": [1.6857613616139133e-07, 2.6294770577806048e-05, 0.9999735355377197], "tokens": 43} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-AR", "prediction": 0, "probabilities": [0.9891389012336731, 0.010851274244487286, 9.751813195180148e-06], "tokens": 41} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-DIFFUSION", "prediction": 1, "probabilities": [0.0011293808929622173, 0.9988576173782349, 1.295601214224007e-05], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-BOTH", "prediction": 2, "probabilities": [9.263276257165387e-10, 8.766419341554865e-05, 0.9999123811721802], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-AR", "prediction": 0, "probabilities": [0.9940400123596191, 0.005960068665444851, 1.3755606376175677e-10], "tokens": 40} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-DIFFUSION", "prediction": 1, "probabilities": [0.018960680812597275, 0.981029748916626, 9.533029697195161e-06], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-BOTH", "prediction": 2, "probabilities": [1.8648794139153324e-05, 2.021743217284211e-09, 0.9999814033508301], "tokens": 35} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-AR", "prediction": 0, "probabilities": [0.9995694756507874, 0.00042987498454749584, 5.602377086688648e-07], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-DIFFUSION", "prediction": 1, "probabilities": [0.00014551721687894315, 0.9998534917831421, 9.282412065658718e-07], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-BOTH", "prediction": 1, "probabilities": [1.4384562985192417e-10, 0.9930213093757629, 0.006978712510317564], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-AR", "prediction": 0, "probabilities": [0.9998987913131714, 9.910601511364803e-05, 2.110677314703935e-06], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-DIFFUSION", "prediction": 1, "probabilities": [0.00456226384267211, 0.9954372048377991, 5.341321411833633e-07], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-BOTH", "prediction": 2, "probabilities": [1.769486956959554e-08, 2.6910389351542108e-05, 0.9999730587005615], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-AR", "prediction": 0, "probabilities": [0.9999964237213135, 3.5193927487853216e-06, 5.360726529346493e-09], "tokens": 40} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-DIFFUSION", "prediction": 1, "probabilities": [0.0007997521315701306, 0.9991984963417053, 1.7247565438083257e-06], "tokens": 37} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-BOTH", "prediction": 2, "probabilities": [9.723816418727438e-09, 4.870445263804868e-05, 0.9999512434005737], "tokens": 39} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-AR", "prediction": 0, "probabilities": [0.9999958276748657, 1.0156242069569998e-06, 3.0637388590548653e-06], "tokens": 41} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999998807907104, 6.85725964899575e-08, 3.71231827500651e-08], "tokens": 38} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-BOTH", "prediction": 2, "probabilities": [2.71955391539791e-09, 5.742134590036585e-07, 0.9999994039535522], "tokens": 37} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-AR", "prediction": 0, "probabilities": [0.9999992847442627, 3.140649766919523e-07, 4.1473779788248066e-07], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-DIFFUSION", "prediction": 1, "probabilities": [1.0106907715259905e-12, 0.999995231628418, 4.739686119137332e-06], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-BOTH", "prediction": 2, "probabilities": [2.30486740804281e-09, 8.219393915487316e-12, 1.0], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-AR", "prediction": 0, "probabilities": [0.9999986886978149, 1.3199968407207052e-06, 4.1519996329952846e-08], "tokens": 37} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-DIFFUSION", "prediction": 1, "probabilities": [1.434372065567402e-11, 0.9999997615814209, 2.6161114874412306e-07], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-BOTH", "prediction": 2, "probabilities": [1.4299308265242061e-11, 1.7693530668383417e-11, 1.0], "tokens": 36} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-AR", "prediction": 0, "probabilities": [1.0, 8.19778467331389e-09, 3.037328388799665e-09], "tokens": 42} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-DIFFUSION", "prediction": 1, "probabilities": [1.1505316876991856e-07, 0.9998125433921814, 0.0001873868895927444], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-BOTH", "prediction": 2, "probabilities": [6.13842743996118e-13, 5.206906086852037e-12, 1.0], "tokens": 39} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-AR", "prediction": 0, "probabilities": [1.0, 1.2264270488060447e-08, 9.650896618321192e-13], "tokens": 43} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-DIFFUSION", "prediction": 1, "probabilities": [4.790859930992042e-10, 0.9999997615814209, 2.7631855914478365e-07], "tokens": 35} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-BOTH", "prediction": 2, "probabilities": [3.5333509122020823e-12, 2.3133719384560436e-09, 1.0], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.2038687202675646e-07, 2.4660046574198624e-11], "tokens": 42} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-DIFFUSION", "prediction": 1, "probabilities": [2.104066053332243e-11, 0.9999998807907104, 9.939152079141422e-08], "tokens": 36} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-BOTH", "prediction": 2, "probabilities": [4.404516196210295e-10, 1.0822828093795533e-12, 1.0], "tokens": 40} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-AR", "prediction": 0, "probabilities": [0.9999997615814209, 4.304290257550747e-10, 1.922962269418349e-07], "tokens": 43} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-DIFFUSION", "prediction": 1, "probabilities": [3.337883303089484e-09, 0.9999963045120239, 3.6863361856376287e-06], "tokens": 39} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-BOTH", "prediction": 2, "probabilities": [3.6023052985001414e-08, 9.178928728204028e-09, 1.0], "tokens": 37} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-AR", "prediction": 0, "probabilities": [0.9851228594779968, 0.013221600092947483, 0.001655503292568028], "tokens": 53} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-DIFFUSION", "prediction": 1, "probabilities": [0.0016196798533201218, 0.9947655200958252, 0.003614793997257948], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-BOTH", "prediction": 2, "probabilities": [1.426634366197277e-10, 0.00015123432967811823, 0.9998487234115601], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-AR", "prediction": 0, "probabilities": [0.9979885816574097, 0.0020114609505981207, 2.061934623043271e-08], "tokens": 52} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-DIFFUSION", "prediction": 1, "probabilities": [0.012553313747048378, 0.9862605929374695, 0.0011861277744174004], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-BOTH", "prediction": 2, "probabilities": [3.637941290435265e-07, 3.721079266938432e-08, 0.9999996423721313], "tokens": 47} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-AR", "prediction": 0, "probabilities": [0.9971627593040466, 0.0027796754147857428, 5.7570894568925723e-05], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-DIFFUSION", "prediction": 0, "probabilities": [0.7629523277282715, 0.23603859543800354, 0.0010091322474181652], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-BOTH", "prediction": 2, "probabilities": [2.2655823883610537e-09, 0.022751476615667343, 0.97724848985672], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-AR", "prediction": 0, "probabilities": [0.9966064691543579, 0.0032343287020921707, 0.0001591928448760882], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-DIFFUSION", "prediction": 0, "probabilities": [0.9670150279998779, 0.032820429652929306, 0.00016448850510641932], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-BOTH", "prediction": 2, "probabilities": [1.4046640517406672e-09, 0.007792472839355469, 0.9922075867652893], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-AR", "prediction": 0, "probabilities": [0.9999992847442627, 6.441098321374739e-07, 1.0917703008317403e-07], "tokens": 52} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-DIFFUSION", "prediction": 1, "probabilities": [0.0025006618816405535, 0.9959775805473328, 0.001521836849860847], "tokens": 49} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-BOTH", "prediction": 2, "probabilities": [8.208371204965204e-10, 0.0012506083585321903, 0.9987493753433228], "tokens": 51} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-AR", "prediction": 0, "probabilities": [0.9999288320541382, 1.5179995898506604e-05, 5.606358899967745e-05], "tokens": 53} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999947547912598, 4.1486309783067554e-06, 1.0747124861154589e-06], "tokens": 50} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-BOTH", "prediction": 2, "probabilities": [1.118511727682403e-09, 9.136409062193707e-05, 0.9999086856842041], "tokens": 49} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-AR", "prediction": 0, "probabilities": [0.9999843835830688, 8.650267773191445e-06, 6.972615210543154e-06], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-DIFFUSION", "prediction": 1, "probabilities": [5.785935532770514e-12, 0.9999983310699463, 1.6488345409015892e-06], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-BOTH", "prediction": 2, "probabilities": [5.557977829084848e-07, 4.355367178021652e-10, 0.9999994039535522], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-AR", "prediction": 0, "probabilities": [0.9999998807907104, 1.0530082761306403e-07, 5.744902775717264e-09], "tokens": 64} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-DIFFUSION", "prediction": 1, "probabilities": [1.1239550262720499e-10, 0.9999994039535522, 6.04382819346938e-07], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-BOTH", "prediction": 2, "probabilities": [2.49666132212667e-09, 1.2940246964243673e-11, 1.0], "tokens": 63} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-AR", "prediction": 0, "probabilities": [1.0, 4.103708928226979e-09, 1.0070070111112273e-08], "tokens": 69} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-DIFFUSION", "prediction": 1, "probabilities": [1.7617617231735494e-06, 0.9999673366546631, 3.088515222771093e-05], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-BOTH", "prediction": 2, "probabilities": [7.247950156719085e-11, 2.006704385992819e-10, 1.0], "tokens": 66} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-AR", "prediction": 0, "probabilities": [1.0, 8.855215227754343e-09, 5.607095950738161e-12], "tokens": 70} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-DIFFUSION", "prediction": 1, "probabilities": [4.757602255267557e-06, 0.9998818635940552, 0.00011337450996506959], "tokens": 62} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-BOTH", "prediction": 2, "probabilities": [7.657939171679118e-09, 1.3999552628263245e-09, 1.0], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-AR", "prediction": 0, "probabilities": [1.0, 1.1039021918790581e-09, 5.911326983465415e-12], "tokens": 69} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-DIFFUSION", "prediction": 1, "probabilities": [2.9960628162434277e-12, 0.9999998807907104, 1.2895011991531646e-07], "tokens": 63} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-BOTH", "prediction": 2, "probabilities": [3.286687189074655e-08, 1.316545397006541e-11, 1.0], "tokens": 67} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-AR", "prediction": 0, "probabilities": [1.0, 9.486307572803554e-11, 4.253224261674404e-08], "tokens": 70} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-DIFFUSION", "prediction": 1, "probabilities": [5.428272276120083e-10, 0.9999984502792358, 1.5707760212535504e-06], "tokens": 66} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-BOTH", "prediction": 2, "probabilities": [5.90521949561662e-07, 4.698600619690296e-09, 0.9999994039535522], "tokens": 64} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-AR", "prediction": 0, "probabilities": [0.9997798800468445, 7.114248001016676e-05, 0.00014896290667820722], "tokens": 46} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-DIFFUSION", "prediction": 1, "probabilities": [4.134959453949705e-06, 0.999955415725708, 4.042940418003127e-05], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-BOTH", "prediction": 2, "probabilities": [5.410352721391121e-10, 6.547214525198797e-06, 0.9999934434890747], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-AR", "prediction": 0, "probabilities": [0.9996615648269653, 0.00033844486461021006, 6.075476566280713e-09], "tokens": 45} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-DIFFUSION", "prediction": 1, "probabilities": [0.0049865953624248505, 0.9942598342895508, 0.0007535244803875685], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-BOTH", "prediction": 2, "probabilities": [1.9242056623625103e-06, 1.842569291587992e-10, 0.9999980926513672], "tokens": 40} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-AR", "prediction": 0, "probabilities": [0.9390603303909302, 0.06080599129199982, 0.00013368838699534535], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-DIFFUSION", "prediction": 1, "probabilities": [1.8633778381627053e-05, 0.9999808073043823, 5.565311198552081e-07], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-BOTH", "prediction": 2, "probabilities": [1.1165010027625044e-09, 0.013392950408160686, 0.9866070747375488], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-AR", "prediction": 0, "probabilities": [0.9998905658721924, 0.00010746611224021763, 1.8785100337481708e-06], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-DIFFUSION", "prediction": 1, "probabilities": [0.05529843270778656, 0.9446864724159241, 1.5108422303455882e-05], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-BOTH", "prediction": 2, "probabilities": [1.8230910114880317e-09, 1.9272545614512637e-05, 0.9999806880950928], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-AR", "prediction": 0, "probabilities": [0.9998794794082642, 0.00012040069850627333, 1.0065193833952435e-07], "tokens": 45} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-DIFFUSION", "prediction": 1, "probabilities": [0.34186244010925293, 0.6576018333435059, 0.0005356905749067664], "tokens": 42} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-BOTH", "prediction": 2, "probabilities": [2.6766797667221454e-09, 2.5996423573815264e-05, 0.9999740123748779], "tokens": 44} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-AR", "prediction": 0, "probabilities": [0.9999772310256958, 2.0653067167586414e-06, 2.07732791750459e-05], "tokens": 46} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-DIFFUSION", "prediction": 0, "probabilities": [0.9999591112136841, 3.6090612411499023e-05, 4.784533757629106e-06], "tokens": 43} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-BOTH", "prediction": 2, "probabilities": [7.593893291968357e-10, 5.705507646780461e-06, 0.9999942779541016], "tokens": 42} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-AR", "prediction": 0, "probabilities": [0.9999619722366333, 3.624451346695423e-05, 1.843613745222683e-06], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-DIFFUSION", "prediction": 1, "probabilities": [1.8995871227996813e-13, 0.9999996423721313, 3.1942673217599804e-07], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-BOTH", "prediction": 2, "probabilities": [2.7655928658276707e-09, 3.503132450788371e-09, 1.0], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-AR", "prediction": 0, "probabilities": [0.9965848922729492, 0.003413622034713626, 1.3809684560328606e-06], "tokens": 46} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-DIFFUSION", "prediction": 1, "probabilities": [3.39357292267195e-11, 0.9999994039535522, 5.574465831159614e-07], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-BOTH", "prediction": 2, "probabilities": [3.7329039165312494e-11, 2.5619493193573817e-09, 1.0], "tokens": 45} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-AR", "prediction": 0, "probabilities": [1.0, 1.829623919036294e-08, 3.4643872215411875e-09], "tokens": 51} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-DIFFUSION", "prediction": 1, "probabilities": [4.03487288025417e-09, 0.9997081160545349, 0.00029180178535170853], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-BOTH", "prediction": 2, "probabilities": [1.8686764375469567e-12, 3.7881919823234966e-10, 1.0], "tokens": 48} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-AR", "prediction": 0, "probabilities": [0.9999985694885254, 1.4214008388080401e-06, 3.865300440830666e-12], "tokens": 52} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-DIFFUSION", "prediction": 1, "probabilities": [1.5420291996429136e-12, 0.9999998807907104, 6.850618206044601e-08], "tokens": 44} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-BOTH", "prediction": 2, "probabilities": [3.937816739352229e-11, 6.802078900136621e-08, 0.9999998807907104], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-AR", "prediction": 0, "probabilities": [0.9999978542327881, 2.170373136323178e-06, 1.512253169910771e-11], "tokens": 51} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-DIFFUSION", "prediction": 1, "probabilities": [4.777925017435125e-12, 0.9999998807907104, 6.074743197359567e-08], "tokens": 45} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-BOTH", "prediction": 2, "probabilities": [1.5691268728801333e-08, 3.1929098254224186e-11, 1.0], "tokens": 49} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-AR", "prediction": 0, "probabilities": [0.9999972581863403, 2.802954979230776e-09, 2.772410653051338e-06], "tokens": 52} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-DIFFUSION", "prediction": 1, "probabilities": [7.088325482129676e-10, 0.9999886751174927, 1.127581072069006e-05], "tokens": 48} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-BOTH", "prediction": 2, "probabilities": [1.350328648896948e-08, 3.194888265056761e-08, 1.0], "tokens": 46} +{"id": "modality-text-global-zh-philosophy/test/165", "prediction": 0, "probabilities": [1.0, 1.5010126475090146e-10, 9.475425999383447e-10], "tokens": 65} +{"id": "modality-text-global-es-human_aging/test/127", "prediction": 0, "probabilities": [1.0, 2.3740910279346217e-11, 1.227828394512187e-09], "tokens": 90} +{"id": "modality-text-global-zh-elementary_mathematics/test/289", "prediction": 0, "probabilities": [0.9999997615814209, 7.520882250311445e-10, 2.447686995310505e-07], "tokens": 75} +{"id": "modality-text-global-ja-professional_law/test/570", "prediction": 0, "probabilities": [1.0, 1.5674126493253615e-12, 7.414500680091862e-10], "tokens": 298} +{"id": "modality-text-global-fr-professional_psychology/test/359", "prediction": 0, "probabilities": [1.0, 3.095582101764194e-08, 2.0211618836918355e-11], "tokens": 51} +{"id": "modality-text-global-fr-logical_fallacies/test/136", "prediction": 0, "probabilities": [1.0, 1.802127457678493e-12, 1.7415473230197875e-10], "tokens": 149} +{"id": "modality-text-global-de-moral_disputes/test/116", "prediction": 0, "probabilities": [1.0, 2.2029880697438387e-10, 7.025792392312269e-09], "tokens": 95} +{"id": "modality-text-global-ja-high_school_physics/test/67", "prediction": 0, "probabilities": [1.0, 5.125765367708368e-10, 4.671421669627529e-12], "tokens": 68} +{"id": "modality-text-global-es-high_school_microeconomics/test/172", "prediction": 0, "probabilities": [0.9999998807907104, 4.888583635120902e-11, 6.638035898731687e-08], "tokens": 52} +{"id": "modality-text-global-fr-high_school_microeconomics/test/123", "prediction": 0, "probabilities": [1.0, 1.6391235591051156e-11, 2.076259650318768e-11], "tokens": 79} +{"id": "modality-text-global-es-professional_law/test/986", "prediction": 0, "probabilities": [1.0, 7.248288883161813e-13, 3.294975403633771e-10], "tokens": 313} +{"id": "modality-text-global-es-prehistory/test/47", "prediction": 0, "probabilities": [0.9984145164489746, 1.3094872031160776e-07, 0.0015853700460866094], "tokens": 100} +{"id": "modality-text-global-zh-professional_medicine/test/56", "prediction": 0, "probabilities": [0.9999994039535522, 4.3501393065525917e-07, 1.2420676398505748e-07], "tokens": 126} +{"id": "modality-text-global-es-miscellaneous/test/639", "prediction": 0, "probabilities": [1.0, 5.483522258487028e-08, 4.7906830502597586e-08], "tokens": 47} +{"id": "modality-text-global-es-business_ethics/test/42", "prediction": 0, "probabilities": [0.9999972581863403, 2.6948910090141e-06, 1.9389737038189025e-10], "tokens": 60} +{"id": "modality-text-global-en-virology/test/107", "prediction": 0, "probabilities": [0.9999985694885254, 1.4538156847265782e-06, 1.1641259137218363e-11], "tokens": 125} +{"id": "modality-text-global-en-moral_scenarios/test/304", "prediction": 0, "probabilities": [1.0, 1.0812513551172742e-08, 6.725058288026275e-10], "tokens": 98} +{"id": "modality-text-global-es-high_school_microeconomics/test/157", "prediction": 0, "probabilities": [1.0, 5.540350400679017e-10, 4.739816872323388e-10], "tokens": 76} +{"id": "modality-text-global-ja-moral_scenarios/test/453", "prediction": 0, "probabilities": [1.0, 1.4132453540760537e-10, 4.569483957883591e-10], "tokens": 123} +{"id": "modality-text-global-de-high_school_statistics/test/185", "prediction": 0, "probabilities": [1.0, 4.0788159516580436e-09, 1.8942053259252134e-08], "tokens": 208} +{"id": "modality-text-global-ja-moral_scenarios/test/73", "prediction": 0, "probabilities": [1.0, 1.4802184478135416e-10, 1.0437765235460361e-10], "tokens": 123} +{"id": "modality-text-global-fr-high_school_macroeconomics/test/270", "prediction": 0, "probabilities": [1.0, 5.5803803418463804e-12, 2.395447729064415e-10], "tokens": 126} +{"id": "modality-text-global-en-professional_law/test/539", "prediction": 0, "probabilities": [1.0, 5.289962478477461e-12, 6.255174156422072e-10], "tokens": 295} +{"id": "modality-text-global-de-high_school_macroeconomics/test/172", "prediction": 0, "probabilities": [1.0, 1.079009859239477e-08, 4.075885939869295e-08], "tokens": 54} +{"id": "modality-text-global-en-professional_law/test/9", "prediction": 0, "probabilities": [1.0, 1.321219939010021e-10, 4.48539694275496e-09], "tokens": 274} +{"id": "modality-text-global-fr-us_foreign_policy/test/28", "prediction": 0, "probabilities": [0.9999997615814209, 7.959900967524636e-12, 2.461277688325936e-07], "tokens": 118} +{"id": "modality-text-global-de-virology/test/107", "prediction": 0, "probabilities": [1.0, 4.164792122285377e-11, 5.834108407286971e-15], "tokens": 157} +{"id": "modality-text-global-ja-moral_disputes/test/331", "prediction": 0, "probabilities": [1.0, 7.228811733234197e-12, 1.2100782598167825e-08], "tokens": 88} +{"id": "modality-text-global-es-college_chemistry/test/55", "prediction": 0, "probabilities": [0.9999980926513672, 1.8873982980949222e-06, 3.7099314731392496e-09], "tokens": 95} +{"id": "modality-text-global-fr-medical_genetics/test/38", "prediction": 0, "probabilities": [1.0, 2.3935489146253985e-09, 1.860454065111261e-12], "tokens": 70} +{"id": "modality-text-global-de-moral_scenarios/test/833", "prediction": 0, "probabilities": [0.9999998807907104, 1.2255483738954354e-07, 7.228018183980112e-10], "tokens": 123} +{"id": "modality-text-global-fr-college_biology/test/28", "prediction": 0, "probabilities": [0.9999997615814209, 9.896512409568459e-08, 1.0204782086020714e-07], "tokens": 92} +{"id": "modality-text-global-de-world_religions/test/23", "prediction": 0, "probabilities": [1.0, 1.071700280602661e-11, 6.83615577340424e-12], "tokens": 57} +{"id": "modality-text-global-fr-international_law/test/92", "prediction": 0, "probabilities": [1.0, 6.862731910528552e-11, 6.361521032172135e-11], "tokens": 99} +{"id": "modality-text-global-en-philosophy/test/52", "prediction": 0, "probabilities": [0.9999969005584717, 2.3395104520318455e-08, 3.138595729978988e-06], "tokens": 53} +{"id": "modality-text-global-zh-high_school_world_history/test/209", "prediction": 0, "probabilities": [1.0, 1.9515240660440691e-13, 2.5345225918016467e-09], "tokens": 473} +{"id": "modality-text-global-fr-philosophy/test/165", "prediction": 0, "probabilities": [0.9999998807907104, 6.761204929262021e-08, 1.6911622102711021e-10], "tokens": 70} +{"id": "modality-text-global-zh-moral_disputes/test/36", "prediction": 0, "probabilities": [1.0, 4.8861494406082784e-08, 3.945019311224485e-11], "tokens": 49} +{"id": "modality-text-global-ja-high_school_physics/test/133", "prediction": 0, "probabilities": [0.9999997615814209, 2.2783142128446343e-07, 2.6035285038972233e-10], "tokens": 69} +{"id": "modality-text-global-de-high_school_statistics/test/145", "prediction": 0, "probabilities": [0.9999995231628418, 6.84464915301497e-11, 5.36018148977746e-07], "tokens": 171} +{"id": "modality-text-global-de-high_school_psychology/test/436", "prediction": 0, "probabilities": [1.0, 5.444860599716772e-14, 7.11539171766723e-13], "tokens": 54} +{"id": "modality-text-global-es-moral_scenarios/test/833", "prediction": 0, "probabilities": [1.0, 4.759847516133675e-10, 1.0684326687271373e-11], "tokens": 142} +{"id": "modality-text-global-en-security_studies/test/107", "prediction": 0, "probabilities": [0.9999997615814209, 2.511999852572444e-08, 1.799431998961154e-07], "tokens": 153} +{"id": "modality-text-global-ja-moral_scenarios/test/848", "prediction": 0, "probabilities": [1.0, 1.9205166623681436e-10, 1.8955080560711934e-10], "tokens": 117} +{"id": "modality-text-global-de-high_school_psychology/test/137", "prediction": 0, "probabilities": [0.9999992847442627, 2.0499215172553598e-11, 6.739946911693551e-07], "tokens": 108} +{"id": "modality-text-global-zh-professional_law/test/789", "prediction": 0, "probabilities": [1.0, 3.601239965955738e-12, 1.7147751968371594e-11], "tokens": 384} +{"id": "modality-text-global-en-nutrition/test/129", "prediction": 0, "probabilities": [1.0, 2.2032421460177476e-12, 1.842905028581754e-08], "tokens": 100} +{"id": "modality-text-global-fr-professional_law/test/378", "prediction": 0, "probabilities": [1.0, 3.106845683498172e-11, 2.162111635561903e-12], "tokens": 293} +{"id": "modality-text-global-ja-miscellaneous/test/728", "prediction": 0, "probabilities": [1.0, 3.0525607042086733e-10, 1.4657203450951783e-08], "tokens": 39} +{"id": "modality-text-global-de-college_mathematics/test/53", "prediction": 0, "probabilities": [0.9999996423721313, 3.3424083767386037e-07, 1.0810633277458237e-08], "tokens": 57} +{"id": "modality-text-global-de-jurisprudence/test/41", "prediction": 0, "probabilities": [1.0, 2.4094864389020542e-15, 2.6303563493979e-11], "tokens": 112} +{"id": "modality-text-global-zh-logical_fallacies/test/152", "prediction": 0, "probabilities": [1.0, 2.130622434037832e-08, 3.020682315391099e-10], "tokens": 40} +{"id": "modality-text-global-en-moral_scenarios/test/56", "prediction": 0, "probabilities": [1.0, 3.220890665289744e-09, 1.0441150027906687e-10], "tokens": 100} +{"id": "modality-text-global-zh-professional_law/test/294", "prediction": 0, "probabilities": [1.0, 1.8417591670510935e-12, 3.38488370665857e-09], "tokens": 80} +{"id": "modality-text-global-ja-security_studies/test/19", "prediction": 0, "probabilities": [1.0, 2.3766100198940876e-11, 2.278554515205977e-12], "tokens": 374} +{"id": "modality-text-global-de-nutrition/test/301", "prediction": 0, "probabilities": [1.0, 8.283704971967509e-11, 1.3784643999176227e-12], "tokens": 94} +{"id": "modality-text-global-fr-marketing/test/84", "prediction": 0, "probabilities": [1.0, 1.3075213654845097e-11, 2.712080782174553e-09], "tokens": 64} +{"id": "modality-text-global-ja-professional_law/test/829", "prediction": 0, "probabilities": [1.0, 8.14271566840219e-12, 1.5341504108756787e-11], "tokens": 216} +{"id": "modality-text-global-es-moral_scenarios/test/51", "prediction": 0, "probabilities": [1.0, 6.554344622600183e-09, 5.619574805493244e-10], "tokens": 148} +{"id": "modality-text-global-es-virology/test/64", "prediction": 0, "probabilities": [1.0, 2.268349258427449e-10, 2.5952380869331826e-11], "tokens": 72} +{"id": "modality-text-global-es-human_sexuality/test/1", "prediction": 0, "probabilities": [1.0, 6.273215835683743e-10, 3.774745793716505e-11], "tokens": 61} +{"id": "modality-text-global-de-virology/test/119", "prediction": 0, "probabilities": [0.999996542930603, 1.2589507036864234e-07, 3.3667197385511827e-06], "tokens": 43} +{"id": "modality-text-global-de-miscellaneous/test/108", "prediction": 0, "probabilities": [1.0, 6.139537211957702e-10, 2.2671997612633277e-09], "tokens": 36} +{"id": "modality-text-global-zh-moral_scenarios/test/628", "prediction": 0, "probabilities": [1.0, 3.358695987998317e-09, 1.5764599348155883e-10], "tokens": 104} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/24", "prediction": 0, "probabilities": [0.9999994039535522, 5.573336281372576e-08, 5.984066433484259e-07], "tokens": 44} +{"id": "modality-text-global-zh-high_school_psychology/test/137", "prediction": 0, "probabilities": [1.0, 1.8468785528691356e-11, 1.2333754462190427e-08], "tokens": 113} +{"id": "modality-text-global-de-professional_law/test/1383", "prediction": 0, "probabilities": [0.9999929666519165, 2.6011109710566416e-08, 7.006630767136812e-06], "tokens": 165} +{"id": "modality-text-global-zh-international_law/test/17", "prediction": 0, "probabilities": [1.0, 3.0930766281578315e-10, 5.3656707549309957e-11], "tokens": 63} +{"id": "modality-text-global-ja-marketing/test/160", "prediction": 0, "probabilities": [1.0, 1.5045641121869124e-10, 9.813862396346185e-12], "tokens": 69} +{"id": "modality-text-global-es-high_school_psychology/test/250", "prediction": 0, "probabilities": [0.9999982118606567, 1.7150872508864268e-06, 1.3742231885771616e-07], "tokens": 66} +{"id": "modality-text-global-de-computer_security/test/86", "prediction": 0, "probabilities": [1.0, 7.219376485512186e-11, 3.4425183515450486e-13], "tokens": 99} +{"id": "modality-text-global-en-miscellaneous/test/251", "prediction": 0, "probabilities": [1.0, 2.3103041257854784e-08, 1.5835524891372188e-08], "tokens": 29} +{"id": "modality-text-global-en-moral_scenarios/test/477", "prediction": 0, "probabilities": [1.0, 3.490393529759217e-09, 7.411231350840097e-11], "tokens": 114} +{"id": "modality-text-global-fr-moral_disputes/test/224", "prediction": 0, "probabilities": [1.0, 5.383646002957154e-13, 1.4696561025442922e-12], "tokens": 55} +{"id": "modality-text-global-zh-professional_accounting/test/54", "prediction": 0, "probabilities": [1.0, 4.115101315260716e-11, 9.729150818316157e-09], "tokens": 83} +{"id": "modality-text-global-fr-professional_law/test/1020", "prediction": 0, "probabilities": [1.0, 3.03520736322227e-10, 8.4659140808796e-10], "tokens": 316} +{"id": "modality-text-global-en-high_school_chemistry/test/88", "prediction": 0, "probabilities": [0.9999998807907104, 1.0943477946057101e-07, 6.44112541081654e-11], "tokens": 82} +{"id": "modality-text-global-fr-moral_scenarios/test/94", "prediction": 0, "probabilities": [1.0, 2.513481334176504e-08, 6.125620899410933e-09], "tokens": 167} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/191", "prediction": 0, "probabilities": [1.0, 1.1533676236297197e-09, 1.0595135879754025e-11], "tokens": 48} +{"id": "modality-text-global-de-high_school_psychology/test/320", "prediction": 0, "probabilities": [0.9999996423721313, 3.4898766898550093e-07, 7.313385452789589e-10], "tokens": 43} +{"id": "modality-text-global-en-elementary_mathematics/test/315", "prediction": 0, "probabilities": [1.0, 1.0156733232236093e-09, 2.339174187682147e-08], "tokens": 44} +{"id": "modality-text-global-de-virology/test/118", "prediction": 0, "probabilities": [1.0, 1.9010240430439307e-12, 2.6021949040000436e-09], "tokens": 68} +{"id": "modality-text-global-fr-nutrition/test/92", "prediction": 0, "probabilities": [0.9999997615814209, 2.06659365176165e-07, 7.0186884082446e-10], "tokens": 65} +{"id": "modality-text-global-zh-computer_security/test/41", "prediction": 0, "probabilities": [1.0, 1.5595780222810163e-09, 1.9041821938681203e-11], "tokens": 51} +{"id": "modality-text-global-en-professional_law/test/1072", "prediction": 0, "probabilities": [1.0, 3.3487584233382606e-11, 1.8027936921072296e-09], "tokens": 326} +{"id": "modality-text-global-fr-business_ethics/test/56", "prediction": 0, "probabilities": [1.0, 3.099446033161257e-08, 4.6979781842537705e-08], "tokens": 113} +{"id": "modality-text-global-en-sociology/test/129", "prediction": 0, "probabilities": [0.9999994039535522, 8.108085758351535e-09, 6.166741854940483e-07], "tokens": 67} +{"id": "modality-text-global-de-miscellaneous/test/413", "prediction": 0, "probabilities": [1.0, 1.6841040784143502e-09, 2.0259102728736877e-10], "tokens": 53} +{"id": "modality-text-global-es-world_religions/test/37", "prediction": 0, "probabilities": [0.9999998807907104, 1.1998845650396106e-07, 1.780741420986942e-08], "tokens": 44} +{"id": "modality-text-global-en-public_relations/test/31", "prediction": 0, "probabilities": [0.99997878074646, 2.1168629245948978e-05, 5.4076682687309585e-08], "tokens": 52} +{"id": "modality-text-global-fr-philosophy/test/6", "prediction": 0, "probabilities": [0.9999955892562866, 4.395772374571338e-12, 4.415751845954219e-06], "tokens": 71} +{"id": "modality-text-global-en-moral_scenarios/test/630", "prediction": 0, "probabilities": [1.0, 1.3727003533858806e-08, 9.389998500974883e-11], "tokens": 120} +{"id": "modality-text-global-es-high_school_government_and_politics/test/135", "prediction": 0, "probabilities": [1.0, 6.046010220250864e-11, 2.2971594071385937e-10], "tokens": 52} +{"id": "modality-text-global-zh-moral_scenarios/test/240", "prediction": 0, "probabilities": [0.9999997615814209, 1.8643694943420996e-07, 4.3129873006364505e-09], "tokens": 112} +{"id": "modality-text-global-fr-us_foreign_policy/test/38", "prediction": 0, "probabilities": [1.0, 9.428574457640004e-13, 3.765038628089634e-11], "tokens": 107} +{"id": "modality-text-global-zh-professional_psychology/test/412", "prediction": 0, "probabilities": [1.0, 2.7183524320406605e-09, 1.210022193554039e-09], "tokens": 58} +{"id": "modality-text-global-fr-moral_scenarios/test/843", "prediction": 0, "probabilities": [1.0, 4.357775651442353e-08, 1.4359677891206957e-08], "tokens": 157} +{"id": "modality-text-global-ja-professional_law/test/1248", "prediction": 0, "probabilities": [1.0, 2.7122976174048796e-13, 4.460304348086197e-10], "tokens": 401} +{"id": "modality-text-global-en-high_school_geography/test/44", "prediction": 0, "probabilities": [0.9743244647979736, 0.025669066235423088, 6.552008017024491e-06], "tokens": 51} +{"id": "modality-text-global-fr-anatomy/test/115", "prediction": 0, "probabilities": [0.9999992847442627, 6.952953413019713e-07, 2.3645606414719744e-13], "tokens": 65} +{"id": "modality-text-global-fr-security_studies/test/238", "prediction": 0, "probabilities": [1.0, 7.44515338269025e-10, 6.884144365670863e-10], "tokens": 348} +{"id": "modality-text-global-zh-high_school_psychology/test/320", "prediction": 0, "probabilities": [0.9999434947967529, 5.65273585380055e-05, 8.211651358891459e-10], "tokens": 40} +{"id": "modality-text-global-zh-moral_scenarios/test/573", "prediction": 0, "probabilities": [1.0, 8.088365532898933e-09, 1.7846160660361932e-10], "tokens": 107} +{"id": "modality-text-global-ja-clinical_knowledge/test/152", "prediction": 0, "probabilities": [1.0, 3.25607385498472e-09, 1.8709878091272003e-08], "tokens": 44} +{"id": "modality-text-global-es-world_religions/test/133", "prediction": 0, "probabilities": [1.0, 1.0477569034605949e-08, 6.228426219223593e-09], "tokens": 36} +{"id": "modality-text-global-de-professional_law/test/479", "prediction": 0, "probabilities": [1.0, 1.81030602834642e-12, 1.8393071787858872e-10], "tokens": 565} +{"id": "modality-text-global-en-high_school_chemistry/test/21", "prediction": 0, "probabilities": [0.999993085861206, 6.93897663950338e-06, 2.8114781613908235e-09], "tokens": 62} +{"id": "modality-text-global-fr-nutrition/test/145", "prediction": 0, "probabilities": [0.9999994039535522, 5.657624910782033e-07, 1.831874929525412e-12], "tokens": 93} +{"id": "modality-text-global-fr-human_sexuality/test/46", "prediction": 0, "probabilities": [0.9999998807907104, 1.6496494481543778e-07, 1.1033036984020583e-10], "tokens": 61} +{"id": "modality-text-global-fr-human_aging/test/29", "prediction": 0, "probabilities": [1.0, 3.220145539106767e-10, 9.302759673701644e-10], "tokens": 62} +{"id": "modality-text-global-en-professional_law/test/479", "prediction": 0, "probabilities": [1.0, 2.4627194381010575e-11, 2.310524571669248e-08], "tokens": 379} +{"id": "modality-text-global-es-high_school_microeconomics/test/123", "prediction": 0, "probabilities": [1.0, 5.673014683883759e-13, 9.500748632262912e-09], "tokens": 68} +{"id": "modality-text-global-ja-miscellaneous/test/624", "prediction": 0, "probabilities": [0.8295010328292847, 6.03395928919781e-05, 0.17043867707252502], "tokens": 44} +{"id": "modality-text-global-fr-moral_scenarios/test/272", "prediction": 0, "probabilities": [1.0, 5.213873421894277e-08, 2.314481051257644e-08], "tokens": 153} +{"id": "modality-text-global-ja-professional_law/test/789", "prediction": 0, "probabilities": [1.0, 5.34749978672866e-13, 6.299128428068068e-13], "tokens": 459} +{"id": "modality-text-global-en-professional_law/test/1020", "prediction": 0, "probabilities": [0.9999996423721313, 8.730010137014688e-08, 2.419193378955242e-07], "tokens": 227} +{"id": "modality-text-global-ja-miscellaneous/test/526", "prediction": 0, "probabilities": [1.0, 1.3226313273773105e-11, 5.447896365799731e-13], "tokens": 42} +{"id": "modality-text-global-en-moral_scenarios/test/622", "prediction": 0, "probabilities": [1.0, 4.693990085513633e-09, 2.2000937530131104e-11], "tokens": 113} +{"id": "modality-text-global-en-virology/test/131", "prediction": 0, "probabilities": [1.0, 1.2902595658559335e-10, 1.2750809956496578e-11], "tokens": 43} +{"id": "modality-text-global-es-clinical_knowledge/test/193", "prediction": 0, "probabilities": [0.9999998807907104, 6.565044685657995e-08, 4.031313238783696e-08], "tokens": 63} +{"id": "modality-text-global-ja-clinical_knowledge/test/32", "prediction": 0, "probabilities": [1.0, 6.153623183929868e-12, 1.8025658743425765e-10], "tokens": 75} +{"id": "modality-text-global-zh-moral_scenarios/test/62", "prediction": 0, "probabilities": [1.0, 1.9500182801834853e-08, 1.796063298087347e-09], "tokens": 112} +{"id": "modality-text-global-es-high_school_psychology/test/265", "prediction": 0, "probabilities": [0.999998927116394, 2.6662433372237615e-10, 1.0430928796267835e-06], "tokens": 63} +{"id": "modality-text-global-fr-electrical_engineering/test/90", "prediction": 0, "probabilities": [1.0, 3.3254388132064605e-10, 2.777432450695727e-10], "tokens": 57} +{"id": "modality-text-global-ja-moral_scenarios/test/180", "prediction": 0, "probabilities": [1.0, 7.996431294898798e-10, 2.2737725591248648e-10], "tokens": 115} +{"id": "modality-text-global-es-high_school_government_and_politics/test/100", "prediction": 0, "probabilities": [1.0, 6.879955996184559e-14, 4.0281868118885455e-11], "tokens": 140} +{"id": "modality-text-global-ja-high_school_microeconomics/test/110", "prediction": 0, "probabilities": [0.999996542930603, 2.7974508611805504e-06, 7.478050179088314e-07], "tokens": 63} +{"id": "modality-text-global-ja-professional_accounting/test/267", "prediction": 0, "probabilities": [1.0, 3.5871763892636466e-10, 5.748803100225075e-10], "tokens": 122} +{"id": "modality-text-global-zh-high_school_biology/test/259", "prediction": 0, "probabilities": [1.0, 7.719749284262889e-10, 3.706202289510685e-10], "tokens": 154} +{"id": "modality-text-global-es-moral_scenarios/test/730", "prediction": 0, "probabilities": [1.0, 4.0518897265195974e-11, 5.26803969370937e-13], "tokens": 143} +{"id": "modality-text-global-fr-public_relations/test/50", "prediction": 0, "probabilities": [1.0, 6.80899170379945e-10, 1.858641018870344e-09], "tokens": 80} +{"id": "modality-text-global-en-professional_medicine/test/69", "prediction": 0, "probabilities": [0.9999827146530151, 1.596656693436671e-05, 1.302846953876724e-06], "tokens": 92} +{"id": "modality-text-global-es-jurisprudence/test/53", "prediction": 0, "probabilities": [1.0, 1.478883682182186e-10, 3.402749637615443e-09], "tokens": 96} +{"id": "modality-text-global-fr-professional_accounting/test/271", "prediction": 0, "probabilities": [1.0, 2.5378818838128758e-11, 3.879159634571222e-13], "tokens": 328} +{"id": "modality-text-global-de-college_biology/test/137", "prediction": 0, "probabilities": [1.0, 3.4997329478869688e-09, 8.821529062785771e-10], "tokens": 58} +{"id": "modality-text-global-fr-high_school_computer_science/test/19", "prediction": 0, "probabilities": [0.9999992847442627, 5.1628206511377187e-14, 7.720955750301073e-07], "tokens": 141} +{"id": "modality-text-global-en-college_medicine/test/10", "prediction": 0, "probabilities": [1.0, 3.779014434712735e-09, 1.6766811272717064e-09], "tokens": 75} +{"id": "modality-text-global-es-human_sexuality/test/70", "prediction": 0, "probabilities": [1.0, 6.725624831366295e-13, 8.916406502024188e-11], "tokens": 125} +{"id": "modality-text-global-ja-moral_disputes/test/196", "prediction": 0, "probabilities": [1.0, 2.4118512742532516e-10, 8.841565500572468e-12], "tokens": 64} +{"id": "modality-text-global-fr-philosophy/test/168", "prediction": 0, "probabilities": [1.0, 7.33596777074963e-13, 6.605013064264753e-11], "tokens": 62} +{"id": "modality-text-global-en-high_school_psychology/test/265", "prediction": 0, "probabilities": [0.9999970197677612, 4.1596454058989707e-10, 2.9234909106889972e-06], "tokens": 45} +{"id": "modality-text-global-de-moral_scenarios/test/767", "prediction": 0, "probabilities": [0.9999995231628418, 4.5644060264748987e-07, 1.8322353412258963e-08], "tokens": 112} +{"id": "modality-text-global-zh-medical_genetics/test/49", "prediction": 0, "probabilities": [1.0, 3.0254350136260655e-09, 3.927419153781919e-12], "tokens": 30} +{"id": "modality-text-global-es-high_school_geography/test/142", "prediction": 0, "probabilities": [0.9999996423721313, 9.974739612061967e-08, 2.7630798626887554e-07], "tokens": 52} +{"id": "modality-text-global-fr-sociology/test/71", "prediction": 0, "probabilities": [1.0, 6.839011995607436e-12, 8.78437678153432e-09], "tokens": 160} +{"id": "modality-text-global-es-moral_scenarios/test/480", "prediction": 0, "probabilities": [1.0, 2.0855532856245418e-09, 6.443153649504652e-11], "tokens": 134} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/100", "prediction": 0, "probabilities": [1.0, 2.0507569253869207e-10, 2.9438417925575777e-08], "tokens": 144} +{"id": "modality-text-global-fr-jurisprudence/test/42", "prediction": 0, "probabilities": [0.9999966621398926, 3.235650410715607e-06, 1.0437069164481727e-07], "tokens": 38} +{"id": "modality-text-global-ja-clinical_knowledge/test/84", "prediction": 0, "probabilities": [1.0, 5.780926085052762e-12, 1.7830575476393898e-14], "tokens": 86} +{"id": "modality-text-global-de-high_school_mathematics/test/45", "prediction": 0, "probabilities": [0.9999998807907104, 1.1614207551247091e-07, 1.089284218380726e-09], "tokens": 88} +{"id": "modality-text-global-en-professional_law/test/1383", "prediction": 0, "probabilities": [1.0, 5.70361924356888e-10, 2.788305941692215e-08], "tokens": 136} +{"id": "modality-text-global-fr-philosophy/test/139", "prediction": 0, "probabilities": [1.0, 3.19592741160335e-10, 2.629638018447622e-08], "tokens": 77} +{"id": "modality-text-global-en-professional_law/test/1189", "prediction": 0, "probabilities": [1.0, 4.3008423211698954e-11, 3.390375757916786e-09], "tokens": 235} +{"id": "modality-text-global-en-marketing/test/90", "prediction": 0, "probabilities": [1.0, 2.207273946952526e-10, 5.587063589551633e-10], "tokens": 59} +{"id": "modality-text-global-zh-sociology/test/14", "prediction": 0, "probabilities": [0.9999998807907104, 1.583855095688147e-11, 6.69709550038533e-08], "tokens": 103} +{"id": "modality-text-global-zh-high_school_government_and_politics/test/58", "prediction": 0, "probabilities": [1.0, 1.0072345735245847e-10, 1.53672751607159e-11], "tokens": 78} +{"id": "modality-text-global-ja-nutrition/test/301", "prediction": 0, "probabilities": [1.0, 1.1970152646867405e-10, 6.121088878818792e-13], "tokens": 114} +{"id": "modality-text-global-fr-high_school_psychology/test/83", "prediction": 0, "probabilities": [0.999997615814209, 2.421190856694011e-06, 7.9245420989138e-12], "tokens": 60} +{"id": "modality-text-global-de-moral_disputes/test/200", "prediction": 0, "probabilities": [1.0, 3.43251815859702e-10, 8.43955838547572e-09], "tokens": 102} +{"id": "modality-text-global-en-philosophy/test/104", "prediction": 0, "probabilities": [0.9999997615814209, 9.317699389832512e-12, 2.970212733544031e-07], "tokens": 74} +{"id": "modality-text-global-en-professional_law/test/774", "prediction": 0, "probabilities": [1.0, 5.035490913130047e-10, 1.09793303337824e-08], "tokens": 359} +{"id": "modality-text-global-zh-virology/test/85", "prediction": 0, "probabilities": [1.0, 6.636678762106385e-09, 1.3126427722554013e-09], "tokens": 52} +{"id": "modality-text-global-en-professional_law/test/261", "prediction": 0, "probabilities": [1.0, 1.2453793551969738e-08, 4.705820733086341e-10], "tokens": 146} +{"id": "modality-text-global-zh-high_school_psychology/test/456", "prediction": 0, "probabilities": [0.9999991655349731, 8.604102958997828e-07, 3.807277604295223e-09], "tokens": 66} +{"id": "modality-text-global-es-moral_scenarios/test/153", "prediction": 0, "probabilities": [1.0, 2.1850772580211952e-10, 2.3697660153643163e-11], "tokens": 172} +{"id": "modality-text-global-en-professional_psychology/test/128", "prediction": 0, "probabilities": [1.0, 7.462017670434307e-09, 4.2612711581568874e-08], "tokens": 64} +{"id": "modality-text-global-fr-miscellaneous/test/58", "prediction": 0, "probabilities": [0.9999998807907104, 3.163646766779493e-08, 9.005053414057329e-08], "tokens": 35} +{"id": "modality-text-global-es-moral_scenarios/test/786", "prediction": 0, "probabilities": [1.0, 5.942817349335883e-10, 6.342887326482582e-11], "tokens": 137} +{"id": "modality-text-global-es-moral_scenarios/test/249", "prediction": 0, "probabilities": [1.0, 3.571474227470617e-10, 1.840748733994424e-11], "tokens": 146} +{"id": "modality-text-global-ja-virology/test/107", "prediction": 0, "probabilities": [1.0, 1.2075904720632025e-08, 2.038630896039617e-11], "tokens": 146} +{"id": "modality-text-global-de-high_school_biology/test/47", "prediction": 0, "probabilities": [1.0, 3.4753446729587564e-14, 9.183682980751229e-11], "tokens": 127} +{"id": "modality-text-global-es-human_sexuality/test/41", "prediction": 0, "probabilities": [1.0, 3.7969807853421855e-10, 5.628099722376767e-11], "tokens": 46} +{"id": "modality-text-global-ja-miscellaneous/test/423", "prediction": 0, "probabilities": [1.0, 1.628635559747238e-09, 4.437537282075965e-10], "tokens": 54} +{"id": "modality-text-global-es-high_school_government_and_politics/test/114", "prediction": 0, "probabilities": [0.9999988079071045, 3.504433010448338e-08, 1.2121269037379534e-06], "tokens": 70} +{"id": "modality-text-global-en-international_law/test/33", "prediction": 0, "probabilities": [1.0, 2.1486132317782847e-10, 2.028664347619724e-09], "tokens": 107} +{"id": "modality-text-global-zh-moral_scenarios/test/155", "prediction": 0, "probabilities": [1.0, 3.5505700601845547e-09, 4.680380527588568e-10], "tokens": 104} +{"id": "modality-text-global-zh-moral_scenarios/test/274", "prediction": 0, "probabilities": [1.0, 7.132241464091749e-09, 3.1373481590435404e-10], "tokens": 128} +{"id": "modality-text-global-de-professional_psychology/test/512", "prediction": 0, "probabilities": [0.9999998807907104, 8.207143054050903e-08, 8.581257926465469e-10], "tokens": 54} +{"id": "modality-text-global-fr-moral_scenarios/test/240", "prediction": 0, "probabilities": [1.0, 5.216429954657542e-08, 1.144626260440873e-08], "tokens": 170} +{"id": "modality-text-global-zh-moral_scenarios/test/281", "prediction": 0, "probabilities": [1.0, 4.161675892788708e-09, 1.3929395414891133e-09], "tokens": 112} +{"id": "modality-text-global-es-professional_psychology/test/80", "prediction": 0, "probabilities": [0.999993085861206, 6.961526196391787e-06, 4.267473874275396e-11], "tokens": 46} +{"id": "modality-text-global-ja-marketing/test/58", "prediction": 0, "probabilities": [1.0, 7.29882196348619e-11, 1.5225005700614247e-09], "tokens": 43} +{"id": "modality-text-global-fr-college_physics/test/77", "prediction": 0, "probabilities": [0.9999998807907104, 7.448426941891739e-08, 4.34462066323249e-08], "tokens": 119} +{"id": "modality-text-global-es-high_school_psychology/test/367", "prediction": 0, "probabilities": [1.0, 3.5206082493743907e-09, 2.4155586419993824e-09], "tokens": 38} +{"id": "modality-text-global-es-high_school_microeconomics/test/153", "prediction": 0, "probabilities": [0.9999802112579346, 1.4274162822403014e-05, 5.439398137241369e-06], "tokens": 56} +{"id": "modality-text-global-en-elementary_mathematics/test/366", "prediction": 0, "probabilities": [0.9996634721755981, 0.0003341187839396298, 2.3987329313968075e-06], "tokens": 48} +{"id": "modality-text-global-es-virology/test/163", "prediction": 0, "probabilities": [1.0, 2.8410896898978244e-11, 6.085014219933682e-13], "tokens": 51} +{"id": "modality-text-global-de-high_school_statistics/test/198", "prediction": 0, "probabilities": [0.9999964237213135, 3.535802989063086e-06, 3.4709293217360937e-09], "tokens": 103} +{"id": "modality-text-global-de-moral_disputes/test/36", "prediction": 0, "probabilities": [1.0, 3.742462695299764e-08, 1.4265768122356803e-08], "tokens": 58} +{"id": "modality-text-global-en-us_foreign_policy/test/77", "prediction": 0, "probabilities": [0.999993085861206, 6.7304042339344505e-09, 6.884977210575016e-06], "tokens": 90} +{"id": "modality-text-global-ja-philosophy/test/39", "prediction": 0, "probabilities": [1.0, 5.740870001602616e-10, 3.55695704001846e-08], "tokens": 72} +{"id": "modality-text-global-fr-philosophy/test/142", "prediction": 0, "probabilities": [0.9999998807907104, 3.3224906159645684e-10, 1.5004481213054532e-07], "tokens": 53} +{"id": "modality-text-global-ja-high_school_chemistry/test/33", "prediction": 0, "probabilities": [0.9999990463256836, 8.121989480969205e-07, 7.848724692394171e-08], "tokens": 68} +{"id": "modality-text-global-de-high_school_biology/test/166", "prediction": 0, "probabilities": [1.0, 6.265272328720428e-11, 1.484450236344248e-11], "tokens": 83} +{"id": "modality-text-global-zh-public_relations/test/66", "prediction": 0, "probabilities": [0.9999996423721313, 3.073766379202425e-07, 1.2200728205513656e-09], "tokens": 67} +{"id": "modality-text-global-ja-moral_scenarios/test/295", "prediction": 0, "probabilities": [1.0, 2.5077335763512565e-09, 1.5043746803833358e-10], "tokens": 126} +{"id": "modality-text-global-zh-professional_psychology/test/472", "prediction": 0, "probabilities": [1.0, 2.952578848081089e-09, 1.8158513581667535e-10], "tokens": 72} +{"id": "modality-text-global-ja-miscellaneous/test/14", "prediction": 0, "probabilities": [0.9950300455093384, 3.8637868016166976e-08, 0.004969931207597256], "tokens": 101} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/274", "prediction": 0, "probabilities": [0.9999985694885254, 1.430453494322137e-06, 1.0813223580430353e-11], "tokens": 48} +{"id": "modality-text-global-de-miscellaneous/test/728", "prediction": 0, "probabilities": [1.0, 6.45414667890809e-12, 2.367899973321208e-11], "tokens": 29} +{"id": "modality-text-global-es-clinical_knowledge/test/80", "prediction": 0, "probabilities": [1.0, 2.1448696985171267e-11, 1.8053915025288692e-13], "tokens": 56} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/110", "prediction": 0, "probabilities": [1.0, 1.2922427572448214e-09, 2.798248611990406e-11], "tokens": 73} +{"id": "modality-text-global-en-college_medicine/test/73", "prediction": 0, "probabilities": [0.9999998807907104, 8.668784801102447e-08, 2.9093641384925206e-10], "tokens": 89} +{"id": "modality-text-global-zh-high_school_biology/test/96", "prediction": 0, "probabilities": [1.0, 3.556946133031341e-14, 1.6730304207984714e-12], "tokens": 78} +{"id": "modality-text-global-fr-sociology/test/188", "prediction": 0, "probabilities": [1.0, 5.392880586541082e-11, 1.4726250441299271e-08], "tokens": 79} +{"id": "modality-text-global-de-moral_scenarios/test/587", "prediction": 0, "probabilities": [0.9999997615814209, 1.8692921344154456e-07, 9.751851770545272e-09], "tokens": 127} +{"id": "modality-text-global-zh-high_school_microeconomics/test/176", "prediction": 0, "probabilities": [1.0, 1.115368561399599e-10, 3.3234812679694414e-09], "tokens": 78} +{"id": "modality-text-global-en-high_school_macroeconomics/test/354", "prediction": 0, "probabilities": [0.9999908208847046, 2.062456161411319e-07, 8.959204933489673e-06], "tokens": 71} +{"id": "modality-text-global-de-miscellaneous/test/365", "prediction": 0, "probabilities": [1.0, 1.2762872356186783e-12, 6.908056904286752e-10], "tokens": 42} +{"id": "modality-text-global-ja-philosophy/test/101", "prediction": 0, "probabilities": [1.0, 1.2616104992299593e-12, 3.105532220271101e-10], "tokens": 101} +{"id": "modality-text-global-de-moral_scenarios/test/111", "prediction": 0, "probabilities": [0.9999997615814209, 2.4637105866531783e-07, 2.8725661849193784e-08], "tokens": 129} +{"id": "modality-text-global-ja-professional_medicine/test/127", "prediction": 0, "probabilities": [1.0, 1.81592310632972e-11, 3.0072805357050925e-11], "tokens": 191} +{"id": "modality-text-global-zh-conceptual_physics/test/10", "prediction": 0, "probabilities": [1.0, 1.6230351507218188e-09, 2.6798960828244844e-09], "tokens": 51} +{"id": "modality-text-global-fr-moral_scenarios/test/148", "prediction": 0, "probabilities": [1.0, 7.534183055213362e-09, 7.276502178577005e-10], "tokens": 174} +{"id": "modality-text-global-ja-high_school_chemistry/test/88", "prediction": 0, "probabilities": [0.9999991655349731, 8.207337032217765e-07, 1.3779776798783505e-11], "tokens": 86} +{"id": "modality-text-global-en-high_school_government_and_politics/test/117", "prediction": 0, "probabilities": [1.0, 2.873657400925822e-08, 5.358305088520865e-08], "tokens": 55} +{"id": "modality-text-global-zh-elementary_mathematics/test/325", "prediction": 0, "probabilities": [1.0, 2.4066013182277857e-08, 7.558522696626824e-09], "tokens": 36} +{"id": "modality-text-global-de-professional_psychology/test/279", "prediction": 0, "probabilities": [1.0, 1.8747366159477252e-10, 9.89019699471072e-11], "tokens": 36} +{"id": "modality-text-global-ja-computer_security/test/86", "prediction": 0, "probabilities": [1.0, 2.2084108153297421e-10, 5.4736932624281295e-14], "tokens": 94} +{"id": "modality-text-global-de-international_law/test/8", "prediction": 0, "probabilities": [1.0, 8.63899576875049e-11, 1.4685908045208862e-09], "tokens": 97} +{"id": "modality-text-global-zh-high_school_psychology/test/510", "prediction": 0, "probabilities": [0.9999924898147583, 7.568409273517318e-06, 3.820083360750459e-09], "tokens": 39} +{"id": "modality-text-global-ja-moral_disputes/test/232", "prediction": 0, "probabilities": [1.0, 1.7842774369114522e-08, 7.150189884619351e-10], "tokens": 41} +{"id": "modality-text-global-ja-nutrition/test/297", "prediction": 0, "probabilities": [1.0, 4.2444818493247993e-11, 1.1990826734309401e-11], "tokens": 136} +{"id": "modality-text-global-zh-high_school_geography/test/110", "prediction": 0, "probabilities": [1.0, 6.78693282130105e-11, 6.698855692377492e-09], "tokens": 50} +{"id": "modality-text-global-ja-nutrition/test/129", "prediction": 0, "probabilities": [1.0, 3.699635270446727e-13, 4.338321021868374e-11], "tokens": 127} +{"id": "modality-text-global-ja-moral_scenarios/test/456", "prediction": 0, "probabilities": [1.0, 1.8610807339669577e-10, 1.1738965355334585e-10], "tokens": 120} +{"id": "modality-text-global-zh-professional_law/test/500", "prediction": 0, "probabilities": [1.0, 1.8498247638526477e-12, 1.6719141626708733e-08], "tokens": 173} +{"id": "modality-text-global-zh-human_aging/test/127", "prediction": 0, "probabilities": [0.9999997615814209, 1.8831035220046033e-07, 7.169248528171579e-11], "tokens": 56} +{"id": "modality-text-global-en-high_school_government_and_politics/test/114", "prediction": 0, "probabilities": [0.9999998807907104, 3.765690337331762e-08, 8.498570736037436e-08], "tokens": 59} +{"id": "modality-text-global-es-high_school_psychology/test/273", "prediction": 0, "probabilities": [1.0, 9.558420678221213e-13, 5.1217950020054914e-11], "tokens": 72} +{"id": "modality-text-global-de-high_school_chemistry/test/124", "prediction": 0, "probabilities": [1.0, 8.746037227780334e-10, 2.2019251699756381e-10], "tokens": 47} +{"id": "modality-text-global-en-moral_disputes/test/273", "prediction": 0, "probabilities": [1.0, 3.121585878673727e-09, 2.382075092732805e-10], "tokens": 53} +{"id": "modality-text-global-de-high_school_macroeconomics/test/284", "prediction": 0, "probabilities": [1.0, 1.2833174523052548e-08, 9.412766573124731e-12], "tokens": 112} +{"id": "modality-text-global-de-high_school_world_history/test/10", "prediction": 0, "probabilities": [1.0, 8.174004008376645e-13, 7.808873448041098e-13], "tokens": 321} +{"id": "modality-text-global-es-conceptual_physics/test/164", "prediction": 0, "probabilities": [0.9998655319213867, 0.00013447090168483555, 3.792882452557933e-09], "tokens": 34} +{"id": "modality-text-global-de-human_aging/test/53", "prediction": 0, "probabilities": [0.9999920129776001, 3.111801305522022e-08, 7.934593668323942e-06], "tokens": 53} +{"id": "modality-text-global-es-security_studies/test/140", "prediction": 0, "probabilities": [1.0, 1.2122379922790971e-12, 2.542420673989909e-08], "tokens": 135} +{"id": "modality-text-global-de-philosophy/test/202", "prediction": 0, "probabilities": [0.9999998807907104, 7.097598597738397e-08, 7.854972672305394e-09], "tokens": 50} +{"id": "modality-text-global-es-professional_law/test/807", "prediction": 0, "probabilities": [1.0, 1.7344298943572888e-14, 1.2145060348037195e-12], "tokens": 308} +{"id": "modality-text-global-es-public_relations/test/61", "prediction": 0, "probabilities": [1.0, 3.801758463595206e-09, 3.298242035043586e-08], "tokens": 56} +{"id": "modality-aya-review-eng-26", "prediction": 0, "probabilities": [0.9999997615814209, 1.1305190383836816e-07, 1.5478734383123083e-07], "tokens": 52} +{"id": "modality-aya-review-spa-18", "prediction": 0, "probabilities": [0.9999163150787354, 7.197820082183171e-07, 8.301819616463035e-05], "tokens": 10} +{"id": "modality-aya-review-spa-32", "prediction": 0, "probabilities": [0.9999998807907104, 1.2453340580975691e-08, 1.3231864670615323e-07], "tokens": 17} +{"id": "modality-aya-review-fra-0", "prediction": 0, "probabilities": [0.9487401843070984, 0.00039615639252588153, 0.050863683223724365], "tokens": 12} +{"id": "modality-aya-review-fra-2", "prediction": 0, "probabilities": [0.9197311997413635, 2.0318449969636276e-05, 0.08024855703115463], "tokens": 22} +{"id": "modality-aya-review-fra-5", "prediction": 0, "probabilities": [1.0, 3.157889594262997e-08, 5.988399220946405e-10], "tokens": 17} +{"id": "modality-aya-review-fra-6", "prediction": 0, "probabilities": [1.0, 1.8841178617279297e-10, 4.5193647435493034e-11], "tokens": 50} +{"id": "modality-aya-review-fra-9", "prediction": 0, "probabilities": [0.9945805668830872, 0.005066430661827326, 0.0003529487003106624], "tokens": 12} +{"id": "modality-aya-review-fra-12", "prediction": 0, "probabilities": [0.9848840236663818, 0.0005439125816337764, 0.014571988955140114], "tokens": 10} +{"id": "modality-aya-review-fra-16", "prediction": 1, "probabilities": [0.41055113077163696, 0.560580849647522, 0.028867987915873528], "tokens": 10} +{"id": "modality-aya-review-fra-17", "prediction": 2, "probabilities": [0.14969231188297272, 0.4058113694190979, 0.44449636340141296], "tokens": 11} +{"id": "modality-aya-review-fra-21", "prediction": 0, "probabilities": [0.9324385523796082, 0.06465860456228256, 0.0029028647113591433], "tokens": 16} +{"id": "modality-aya-review-fra-23", "prediction": 0, "probabilities": [0.999927282333374, 6.38879428151995e-05, 8.78763876244193e-06], "tokens": 10} +{"id": "modality-aya-review-fra-24", "prediction": 0, "probabilities": [0.9999990463256836, 7.285318019967235e-07, 2.8770097060260014e-07], "tokens": 17} +{"id": "modality-aya-review-fra-25", "prediction": 0, "probabilities": [0.9994121789932251, 2.6937788888403702e-08, 0.0005878446390852332], "tokens": 14} +{"id": "modality-aya-review-fra-26", "prediction": 0, "probabilities": [1.0, 1.0962897256661108e-09, 3.04242186999204e-11], "tokens": 27} +{"id": "modality-aya-review-fra-31", "prediction": 0, "probabilities": [1.0, 2.4457776248709706e-08, 1.5891134979087451e-12], "tokens": 56} +{"id": "modality-aya-review-fra-33", "prediction": 0, "probabilities": [0.5646347999572754, 0.00016770132060628384, 0.43519753217697144], "tokens": 22} +{"id": "modality-aya-review-deu-19", "prediction": 2, "probabilities": [0.3285082280635834, 0.0007068871054798365, 0.6707848906517029], "tokens": 24} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:head", "prediction": 0, "probabilities": [0.9944714307785034, 6.5602716858848e-07, 0.005527916364371777], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:middle", "prediction": 0, "probabilities": [0.9981608986854553, 8.028739102883264e-06, 0.0018310953164473176], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:tail", "prediction": 0, "probabilities": [0.9984147548675537, 5.987766053294763e-05, 0.0015254024183377624], "tokens": 4096} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:head", "prediction": 0, "probabilities": [0.6995566487312317, 0.00018061348237097263, 0.30026277899742126], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:middle", "prediction": 0, "probabilities": [0.8464436531066895, 0.008684688247740269, 0.14487165212631226], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:tail", "prediction": 0, "probabilities": [0.792873740196228, 0.003597539383918047, 0.2035287320613861], "tokens": 8192} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:head", "prediction": 0, "probabilities": [0.5019932389259338, 0.000340887694619596, 0.4976658821105957], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:middle", "prediction": 0, "probabilities": [0.47082823514938354, 0.07442700862884521, 0.45474472641944885], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:tail", "prediction": 2, "probabilities": [0.18649403750896454, 0.00015124809578992426, 0.8133547306060791], "tokens": 16384} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:head", "prediction": 2, "probabilities": [0.01714290864765644, 0.0019352142699062824, 0.9809219241142273], "tokens": 32768} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:middle", "prediction": 2, "probabilities": [0.012050176970660686, 0.009946445934474468, 0.9780033826828003], "tokens": 32768} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:tail", "prediction": 2, "probabilities": [0.0717330276966095, 0.0065806275233626366, 0.9216862916946411], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:head", "prediction": 2, "probabilities": [7.236756971451541e-08, 5.868127573194215e-06, 0.9999940395355225], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:middle", "prediction": 2, "probabilities": [2.679126964721945e-07, 0.00018964811170008034, 0.9998101592063904], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:tail", "prediction": 2, "probabilities": [3.682347582412149e-08, 2.0979174223612063e-05, 0.9999790191650391], "tokens": 4096} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:head", "prediction": 2, "probabilities": [5.86796318202687e-07, 5.282657639327226e-06, 0.999994158744812], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:middle", "prediction": 2, "probabilities": [9.865346584092549e-08, 2.6266103304806165e-05, 0.9999736547470093], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:tail", "prediction": 2, "probabilities": [9.850497306729267e-09, 6.919846055097878e-05, 0.999930739402771], "tokens": 8192} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:head", "prediction": 2, "probabilities": [5.780439096270129e-07, 1.1744235052901786e-05, 0.9999876022338867], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:middle", "prediction": 2, "probabilities": [4.770511736751359e-07, 0.0003399146953597665, 0.9996596574783325], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:tail", "prediction": 2, "probabilities": [2.1067248212602863e-07, 0.001980509841814637, 0.9980192184448242], "tokens": 16384} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:head", "prediction": 2, "probabilities": [4.690979949373286e-06, 8.143830200424418e-05, 0.9999139308929443], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:middle", "prediction": 2, "probabilities": [1.8093037112976162e-07, 0.000358739634975791, 0.9996410608291626], "tokens": 32768} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:tail", "prediction": 2, "probabilities": [9.915629561874084e-06, 0.030556028708815575, 0.9694340825080872], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:head", "prediction": 1, "probabilities": [1.8590350009617396e-05, 0.9962459206581116, 0.0037354768719524145], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:middle", "prediction": 1, "probabilities": [3.1652632515033474e-07, 0.999873161315918, 0.00012652110308408737], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:tail", "prediction": 1, "probabilities": [1.2182235877844505e-05, 0.9981144666671753, 0.0018733206670731306], "tokens": 4096} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:head", "prediction": 1, "probabilities": [0.00048283845535479486, 0.6680790185928345, 0.3314381539821625], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:middle", "prediction": 1, "probabilities": [5.309087282512337e-05, 0.9783055186271667, 0.021641341969370842], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:tail", "prediction": 1, "probabilities": [2.66728443421016e-06, 0.9994633793830872, 0.0005340005736798048], "tokens": 8192} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:head", "prediction": 1, "probabilities": [0.0001845526130637154, 0.7088466882705688, 0.29096874594688416], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:middle", "prediction": 1, "probabilities": [0.00036718047340400517, 0.7995266318321228, 0.2001061886548996], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:tail", "prediction": 1, "probabilities": [1.6829943660923163e-07, 0.9993209838867188, 0.0006789052858948708], "tokens": 16384} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:head", "prediction": 1, "probabilities": [0.00013220793334767222, 0.7062201499938965, 0.2936476171016693], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:middle", "prediction": 1, "probabilities": [0.00019778271962422878, 0.8385413289070129, 0.1612609177827835], "tokens": 32768} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:tail", "prediction": 1, "probabilities": [0.0004870798729825765, 0.7612209916114807, 0.2382919043302536], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:head", "prediction": 0, "probabilities": [0.9999994039535522, 8.347871016667341e-08, 4.311854979732743e-07], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:middle", "prediction": 0, "probabilities": [0.999996542930603, 3.18887487082975e-06, 2.391612667906884e-07], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:tail", "prediction": 0, "probabilities": [0.9999788999557495, 1.0508685591048561e-05, 1.0554967957432382e-05], "tokens": 4096} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:head", "prediction": 0, "probabilities": [0.9999736547470093, 1.5008223272161558e-05, 1.1297664968878962e-05], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:middle", "prediction": 0, "probabilities": [0.9999668598175049, 2.1992449546814896e-05, 1.1205511327716522e-05], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:tail", "prediction": 0, "probabilities": [0.9823050498962402, 0.017543990164995193, 0.000150909589137882], "tokens": 8192} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:head", "prediction": 0, "probabilities": [0.999961256980896, 1.8482591258361936e-05, 2.029981442319695e-05], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:middle", "prediction": 0, "probabilities": [0.9788746237754822, 0.00915475469082594, 0.011970596387982368], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:tail", "prediction": 1, "probabilities": [0.433368980884552, 0.5651097893714905, 0.0015212201979011297], "tokens": 16384} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:head", "prediction": 0, "probabilities": [0.9980225563049316, 0.00010695246601244435, 0.0018704839749261737], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:middle", "prediction": 1, "probabilities": [0.41942644119262695, 0.5757708549499512, 0.004802683833986521], "tokens": 32768} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:tail", "prediction": 0, "probabilities": [0.981913149356842, 0.0032700656447559595, 0.014816692098975182], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:head", "prediction": 2, "probabilities": [3.52830511474167e-06, 7.340169361214066e-09, 0.9999964237213135], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:middle", "prediction": 2, "probabilities": [1.1979438568232581e-05, 2.5684603333075984e-09, 0.9999880790710449], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:tail", "prediction": 2, "probabilities": [8.985964996099938e-06, 4.03763387168965e-08, 0.9999910593032837], "tokens": 4096} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:head", "prediction": 2, "probabilities": [0.0010569399455562234, 1.5419546528505634e-08, 0.998943030834198], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:middle", "prediction": 2, "probabilities": [0.00014552928041666746, 6.893341453206858e-09, 0.9998544454574585], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:tail", "prediction": 2, "probabilities": [4.841924692300381e-07, 1.7501827542787396e-08, 0.9999995231628418], "tokens": 8192} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:head", "prediction": 2, "probabilities": [0.0018502868479117751, 2.3921810310412184e-08, 0.9981497526168823], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:middle", "prediction": 2, "probabilities": [0.12794692814350128, 2.0167915337765407e-08, 0.8720530867576599], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:tail", "prediction": 2, "probabilities": [0.00020223566389176995, 4.278267340396269e-08, 0.9997977614402771], "tokens": 16384} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:head", "prediction": 2, "probabilities": [0.0007165094139054418, 7.385167322127018e-08, 0.9992833733558655], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:middle", "prediction": 2, "probabilities": [0.01032492145895958, 2.915413244153342e-09, 0.9896751046180725], "tokens": 32768} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:tail", "prediction": 2, "probabilities": [0.0028887521475553513, 1.3171639068332297e-07, 0.9971110820770264], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:head", "prediction": 0, "probabilities": [0.8946110010147095, 4.399930185172707e-05, 0.10534495115280151], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:middle", "prediction": 0, "probabilities": [0.8322411775588989, 0.0007996667991392314, 0.1669592261314392], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:tail", "prediction": 0, "probabilities": [0.9988127946853638, 0.00010989873408107087, 0.0010773605899885297], "tokens": 4096} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:head", "prediction": 2, "probabilities": [0.08928164839744568, 7.222004205686972e-05, 0.9106461405754089], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:middle", "prediction": 2, "probabilities": [0.13784384727478027, 0.00021293375175446272, 0.8619431853294373], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:tail", "prediction": 2, "probabilities": [0.28324827551841736, 0.003489355556666851, 0.7132623791694641], "tokens": 8192} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:head", "prediction": 2, "probabilities": [0.032626599073410034, 0.00033066156902350485, 0.9670428037643433], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:middle", "prediction": 2, "probabilities": [0.010822119191288948, 0.00288560101762414, 0.9862922430038452], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:tail", "prediction": 2, "probabilities": [0.04243037477135658, 2.2737110612069955e-06, 0.9575673341751099], "tokens": 16384} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:head", "prediction": 2, "probabilities": [0.49186909198760986, 0.000529891112819314, 0.5076010227203369], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:middle", "prediction": 2, "probabilities": [0.06488325446844101, 2.9391403586487286e-05, 0.935087263584137], "tokens": 32768} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:tail", "prediction": 2, "probabilities": [0.14395827054977417, 6.142452912172303e-05, 0.8559803366661072], "tokens": 32768} diff --git a/evaluation/final-evaluation-plan.json b/evaluation/final-evaluation-plan.json new file mode 100644 index 0000000000000000000000000000000000000000..d0ea9dfcf37647e9f2a5466fff8158316d9112f6 --- /dev/null +++ b/evaluation/final-evaluation-plan.json @@ -0,0 +1,13 @@ +{ + "candidate_lock_sha256": "34c6015eb91e0fd500371f9bad66cf28600533acb0c6197022eff3b3eeaa3b5f", + "candidate_frozen_before_evaluation": true, + "test_selection": false, + "precision": "FP32 parameters and FP32 forward", + "baseline_revision": "9f05f15889d32987d29bbe2e71bff945902ff553", + "baseline_pooling": "mean", + "candidate_pooling": "mean", + "timing_is_service_latency": false, + "natural_labels": "Aya requests individually assistant-reviewed; academic text-answer and original output-contract templates are not naturally observed three-way routing requests", + "historical_pretraining_or_old_task_exposure": "Unknown for original weights; split isolation established for this new repair training only", + "long_context_source": "Authored stress with frozen held-out payload/source groups and backgrounds; repeated prose is not a natural long-document benchmark" +} diff --git a/evaluation/final-label-manifest.jsonl b/evaluation/final-label-manifest.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..f01c09200a62133011709af430ff0df79c5732d8 --- /dev/null +++ b/evaluation/final-label-manifest.jsonl @@ -0,0 +1,1411 @@ +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ee45f320ba901f392f1f7a6a4f3bf69af5ac575bd42a0e8473110030425185c9"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3f092b40ed4b9d622a07e0aa087715d1e99548cb72691aa03b27a3ecac778f4a"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-en-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "eee99d2cb4dc30cf750b529e0fbdcbf8e1aad17fc116b5dcec0e6222c67aae5f"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "59fc9d233e1b508bd23967a9a70324a63fe69e3cc86410431f57da95512ecf84"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "97f933d72092cf2b374bf642e958ff28ea542f4100c32b6fc4ed7a8b201b3b03"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-zh-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "4274500d22a4c4419670e9099540753cd50b8cfa716543d1b4af3f7bcfd1fc41"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "2c9a06f5e8863c73dc964ac58fef6f5e446bdb7cd97cf5ee7df0537b6cbe6a6b"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1920aad81ab8ccb372f490b6298e9bbad3106b5d12d11b590dfd9d161e52f73e"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-es-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c644bf988a5c9d1f3a922db64f4cf93537402a5498c76a3697be9b9368f3ff69"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a8fc521d57723ab710da47812439121ccca0bd401ef1ad08d01986323e7791b7"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f7836f85e28d99914fa6670c1e467ffc7947cc0df0e7a6d6c71145303a8431ba"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-fr-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b44dde0cf3fab32baf0c58b5979311f47e26574e041c35764058bd7c1306c3ee"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "47b0c152992c045672f081049447e0b8c8d830208f342113758f38c38c0ceac0"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4040c52d7806a2c2645a4246f1f333d611feaeb70e4093037a2b8a1503dd4706"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-de-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9fafb94c9566586937ba2afb0414003d06980bd0329d4ac9ab44977c124bc5ad"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-AR", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "c6a2e7517e44dc256350006539990e5d3ab684aa037bbf1d0294e5face99363f"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-DIFFUSION", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8cb798ea53ba3c0d07c04696ed68afd193c92abf2fe7f05be8094b397246cb78"} +{"id": "modality-00000f6f9849da94c231c1c875469790e759d58d2385f8b708df73081496582a-ja-BOTH", "group_id": "diffusiondb-author-2806a3c0544be83ef38a3e1be82a36e2bd79201eebb0257252fdfbe6e55e1901", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ebabc669635e3c30d99edddaabe777cc27988386fa854ecd68bff158a584297e"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "1c255c4eb3421ca056ab7ddb2e0f23626639489904fb7ee6f2cc4c0b3121966b"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ec644b05892b013d5b3f94b1536eba5f5125c44fbd1fb47ced006afad1dab808"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-en-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "0b90266f71bf77f1798454ba2a98af471ace8b952cf4a26797bbff4d6eba8377"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0ed529446fb25b9912d355600531e62745ef226d3897d7862ded713997cba310"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "af3a1a426c6a8e9b946500caba613a77abe4c89b79000fe158586f8873db99b4"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-zh-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "58a51c51bc43f63ebf46d69c52fbad096c50986ca3435d12beafb113c33875d6"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4e6b22a33b81d870a003d299b617d3d417da6e9ca748f1a542b23b35e5775a38"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1448b1de40410f358caa75820aa15cf4c3fdc4df092ddbba37d15335193b9d2c"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-es-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "2de285ebcf5596866e38f9ab3d45cb8c7bd9927a70d44c61a731b9c0f3ce63aa"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "aa17507204ef339fa21cb539f1a5fb0a6ab04969468689d2412bdb10e07c7be3"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "2bd8a3b64b72074687503e587d7576e1701334fadde02c86e399946d36d651a5"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-fr-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a0d49daff590a23035f576cf7482a34c2c14c231d147ed41ed953db29b367cfc"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "bfdddd68b26ca941679d912a4d0248d5ca2dd7c59cad087b1869ec60643f996c"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "bac7ab3598cd7afd782d5c667ef6a502ee375a7220a7f581b0608ff0f56e6d19"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-de-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "fe301a6cdf8644dfc95df34166c8153823952d071e65f4cc4c274ba4f866e7c6"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-AR", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "cadb42ed086fae483fa2110bd91b14e51a1c097eceea0ae5af1dd3474e3618aa"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-DIFFUSION", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1145bd1b280168758fc8e2b5d80a49e5110aad91148736135bb42609cbb553d5"} +{"id": "modality-000065445bd6f53a8342472700c4ba65bc17ac395356fb87fac8fa50daac79b6-ja-BOTH", "group_id": "diffusiondb-author-ca524b99e00a1f5e66bc48c58e9c9300019d94c85d82389954caa5221b41f786", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6ed6cacdc1c5f1f7fcdc4750b511e46745cb30d5c6a2e41de1c0f8f877415963"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "9a104d6b8f2f2b942ab454f9f30d65f31db1f8caa1bd93feee05477beca68a45"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "17f973bc659b558b7870522bbf6701412d7f1404c21313e53a99951949b26db8"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-en-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "c3390d1fa9dc89e3ebb43cdfae9944908433a217858de4bc50281c80cec30c3f"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1307c776bdd17090b52c2a0571183fc08cb3613d40f1781f07366e888ca79c99"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "889ae0ba28cc72cc16a5517283068766ac49b90ae6ad2b0b26d1413347e0afb3"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b0cc6fb849f7e2250b965ef5b2dee682ad1e922aae33cb231677c6179b63910d"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e55c2ebe14116f4db65f5adc6807ccc55c95622c8bee4a9dc903704f0e442e04"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9111c9ea308f268f8d3d479995b1db3ae1279d092b54415b48eba462c34dbc25"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-es-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "b8020de4cefcc2029c9e2e0ed6861172a814037061b73385b358275fc7e4c51a"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e868de3aef9307bc7b523f737f51e121a6267396cdc220f6e1d4d66a94ed7644"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "36cc04abe39c5cae7050c2a4fadc49e79b734cd834dd066c634d484f8e29cb3b"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-fr-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6370cae29d2d71297d9bbf5b880033b2f7db92008b8958c76aa799e8a1679bd5"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6257075a38bc3fcd4fc2cec1cec22a192fe670efdbf436a98a1c1bb35d437bc6"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e5e08f0253f425df2bcdd3985e9f9fffee7a0ea4aeffbd2d73453cd190586d29"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-de-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "475e4c2002add7dc110e1f397bf8d4cda479a53ca2b01e4590c9ddc8a9e0e4f2"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-AR", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b5bc04c3f41ffd860484564a6a186656ed86575dae3ca97d220ed2cf8bbb279f"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-DIFFUSION", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d6a620b9424d4b4a1505e1411672e554960353b4848670972dac6c5f852f6878"} +{"id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-ja-BOTH", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "2c73097341443cdaf4e4b3e810dbad9973407b1c6f0a08e0326eb383755bc661"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "089c90a7a4add0c8c00b2d450759a16e2d95a5876432e12cf5ed4f159e410be7"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b43c56d5249aa5dd8f5c4f25cf6bf7ecfd2699759deecb02b087b20ec7433974"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-en-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "be4068b44ef241f6bc20438ce6825be8a2c799a4cef0538449f7b307c0b99050"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8f9beb2587fc079939b5e00dd5b2e3824a2c8ab083fe6a8616da7bf15870cad6"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2f57558df93a49fa3ff36941da866d9d06339fb78cfecd9f58a4c8ae2eaebca1"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-zh-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "9e91973db79a7c9d3633a26b546328fa41ced789e2b0b31a59aa3f2c15ca496c"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a02c1792c7a82f9faa54a9b765611962c06766f952c28207d64f9fded93a4e33"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e609ecfd3edf068f2941ffe9c9a7eaf1f09910183aa9ed3ff0a5704d1d3c7396"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-es-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "88c9f622467c6d39727ac186651dee580702f89a7d31bfe7abad485907ac0c05"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "17959487b2aa8dbe816f6135531da0b7e4b75e75418889cf884f6aef25c3dbe3"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "11b4d75274be1729b0fe9edc1ca17d0c63c54e28e5fde041756e90a28222e846"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-fr-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "475edc4615ff0b04d2e6e6925553c26d0494bb1202bd6350f8d82447b62b9575"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "559360d42153a3e7d4b97431bdbfefbdd3ee985258304b62a9d226687507d757"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "b812672291e25bc2fe9013c0839d3c76ad6fd8b5ba32a4b041cdfd178dabfd23"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-de-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "222301e0ece76d49e214232fc307ca5df127c7b4c9ccabb777ab8f6060602e61"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-AR", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "a246621b66d22982f84d5a3f5f65e314ad6854d4695ab14ad94ceae9c51e6131"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-DIFFUSION", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "854ed2e1fb2ca446e28a0e03a62433427886c49f1f9732e5f1b44063995d2d74"} +{"id": "modality-0000a429336db5923e3b481431fb3bf3bc1aba113365b71633b2a87c1cd50387-ja-BOTH", "group_id": "diffusiondb-author-1f9e6243b19daf9945c135f2c4fbb7f95207426f7ecaee9b40c1287c2f679d45", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "c1f60984de904db741657f8d53aecf4b449978407a9afb97245d6f5e13eb48e9"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3fb6caf5e51ef54c65a1ae37960420ebdd00e70eb43338a18343ea05a4b98ead"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e3c0676ae425a06e093ea7c39014279685bd0956f65cc882a0d0f17ad6885064"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-en-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "fa929cad6a294e63d92bf8de3b52a886b79b5e3432673ac9e2a36cde5a87a181"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "fdb765c7d10fa02527c71e5f936750e1af8f3326f267029383c997cc7beaf067"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "6bcabeacc0fc9b701dbf1885e9b16e61470aed729935b37bf3ca5b6b3f2fe6d8"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-zh-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "725f5ae2d2befafc6f93324159e9bc7623ec61b5dc0658df72b2bb143514af7e"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9f855b62743d3e5e9f756e4f190ffb52c6ecec7e46bf36a18827ef28c07ca932"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "7646f584c4a8ba67aa4f4782c45df60f0c17ab9a50e7f76bb767f961d928d6e1"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-es-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "41cc84ee912a16bdd105645a56172b605883a3794a3b25fa178dee5d45d6a943"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "dee99bf177386c3103f1c6164ac694090e4f3a9416f88bc4164762314aba371f"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "551953545fa2973b7c8724ed3b668c69756e95b80b5216c2d0c525dc4a4cf3a7"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-fr-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "2b6e9ae2bb2e623d0d4fc56ad3299d1c8b8b61d10cf0c24206f0c3a5b061f927"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6b47af5a394c8a8c0b8bd0077b8bc2c0142cbba56f24206afe16bb5b656543be"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "09b6bc5bf5482d57c930b24667ea191db8bb99c01ee5c78c8dcc989ed8c1f486"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-de-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "fe5fb16056f541bd23e3d17ebb49488711fdf34636b2635191fa6d1f068220ad"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4457e8767443761d3ea0539ce8d36848ac838847b752dbff72e96842929398dd"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6370c9ca7616f3bc100567ed3802aff2c910febfded9b8663073c3d6482dcda1"} +{"id": "modality-0000b0f302a52e46e1ce84e4e37edc5f2a478b8c5115325ca9ebcb099afa4eed-ja-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d71795c2e39e1cb0d1e73d70d6603940b343d53fcf1432b7d8e24d7f579bf393"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "23f472193404f75788c845acefaf453c339a21a51de96b964cf5a7801a37b445"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "04c04a267f8ea6ff2bd38521e4e54c3293f37972bd77da769db10ea9375f690d"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-en-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "2ac2043a621d5474be3a76e3dda713516cef16c34d6370ded6b02b4994f1f1b5"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "cede8385a0bf9523909992f858ddb884c5f8ce4379f6186a164a32162c2e459b"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c20de2bbf60d9e024ce9e1c1d60dd161b11a5ca712c6c789f91af3c5e41ac9a0"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-zh-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "5195d2fb0db093c71798b088ee4a743a5ae72315d0f16b644b24f8b40b0428c5"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "39dc7f5eafce5ce2016e83821fcf0561c49725621830bcd18b214ed2989cac6b"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "bf5f7ebd6079a8ec214e6a67e2aaddd04cbce2dade04ce4fcabfdb3bd0822e03"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-es-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "26d98e49dbfc0a17ce10f15dfbb87b0ee7879b7a5dc07222198588a490d0eea1"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0524c10519767da0c5d0c7ece53857b152854816859fae4a1063c2006bbe4a1a"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "886f235eab6e520ec4a7b3984bdaa2cf7eecf3e485794cdbd7611f673e1f50df"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-fr-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6b0f832e05c94dee6b8908626875f0d6f80e152b9744035c276df5eada2c350e"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "bdbb78e8bc6349cc255f55c2bfbe401fb0d82bc42f88c8cf585fd0b76f8e3511"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f74d78a035f7bcd1fe05b03dd2d867bf6b44abb2c9a5c528f34198267dbfcfa3"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-de-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "880aadc424824a80b28c2e3326f6c950c1843b365fd32f17876d3b5d48f466e0"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-AR", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "31e4d8829fd625fee865b3e9efc7a9fb91c4f14969984325523727bbfa69bf89"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-DIFFUSION", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d56c520506539bd92534ace17916fc4c17f6df7bdd67cbdc96439495ef104192"} +{"id": "modality-0000e2e6c2890a548a4411d7ebddafe8ab707dedbf91ab5d19641e0a00a64735-ja-BOTH", "group_id": "diffusiondb-author-0727a72676ff915f6ae93e2ebe0b465afedfec537ddb3bbbf162117ee5f4d785", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f6ef3f64eb5230946a090356d5e03a85ca83c59f7f72267ca7fb2f76b81dbbf6"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "c46559f44f428625d08f4cb71474524e38bca304edb27781ac7fc921719a07af"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b90590a42f36dcbaa1f37f4053063ac74eee6bc981ede7137de6103dd0075fa9"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-en-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "0eab8dc90064f79cc217e722bfceb929b94b9e105824f2a94781fb49d7311f67"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "6475a07f8fd6feb0fc438e9111ddba8fcd40c3a27f080b3fdd44d7c6e42b10b0"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8b32eb5e71584c79bc05e3c19b6f3558f83af216cc3de1d07e61d9977812b7be"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-zh-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "81d01417b2b3537e36d6308fcb7ae7d8d9273b14efa2298732c010703a72c3cd"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1106df96128764b512b0a7a5194467ba1f0b5f95a51a502fee0ed870c51e728d"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a0f963430672cf78a72e942a756c77733041606b8ef13f31bcf6614fc243e57e"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-es-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "7d357e17e98c803eaee9a7194bab11720fefd1d6af54fcd3c56e096eabb13674"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "299465b039b4520fee8c6cf61fd6f76bc99e6be319a878975a6a65aaa3d5a5fc"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "15c302e692984af5e068a55726b584fa74cf73af172c323fd4a950c654643863"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-fr-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a3c868e01b87a7d06fbe7268a6ed03b8ac44e7c06e69e85cc8f648f5aa3313d4"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "95d02392786f82d84a232712a49041712cad5ce93d3e11912190113b7d9c0ee2"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "61754da4edeb993d846be2fe3e96866d42050304df0442b262e3deffc7bc8213"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-de-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "27e3b165e2517d2e28d4b3c73941df27e5fd2270e98e07a89d1d8801b970845c"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-AR", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8d5e097d5ab1ba064a0a4fea0d1fb6fa55e686056a16ee3b580b4fa68b3d03d4"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-DIFFUSION", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "bad33547273057104c964e98117e1fe030593d8201d398e2f13628237cf346c7"} +{"id": "modality-0000e9fb7c618249dfab9ed5b76a233efae393efa4e0fdb375271e796047204e-ja-BOTH", "group_id": "diffusiondb-author-c9ebc940eb8a2d7428e617f3ccac38c3c8a00466cda2a19daca21419c1d845b8", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "e9a0d6e476dac95337e24b65c3c13128881c109404121c66c135e023719495f0"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d3beda299d543289253225015b30ef27e7731483cb01639017b3436c3bcee344"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "f2c5ebec770c5566a1dfce822bfecafb3a0be29d7afd7ea3b319c30688ab5c73"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-en-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "30fc92db4ebc830c5d93bc5733dbc5e49b969b9bb6f09e55490be3cd42d40600"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b94ce786d9f4563bf6c6937177efb02dae57f8cd3f312b81c587c11670936bcd"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "ade2d85ea0776e647b71bca8e8128ec39b5e639b307915c77e24723b5f2d8e6d"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-zh-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b03ec0e9bc90c1eddaa1098a7ed41f253fbf09edce8f47be03e30fde7c013e38"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "5263e705b09df0bc1cc5455e857368e1a74b66ce5fc8261fe15e9bf511d36386"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "81fc733dd0c8207f5c6d9338ec61f4cf032791a1c4cc9308744d4e58f0fe64e9"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-es-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ff68ccf73d6c51540ee798a900439d3db221441b1b83e7a8d280623028a523fd"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "8ab0c3ec802029a58214f4976d5221fcbb963b5c620e82177091cda661a363b5"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "49fc261e34398bd23d4116365f4d538c446f2a26079cca320b5f39e17a4ffe25"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-fr-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5914e1ff303fdcc9f2dd6a67d1efff59fcf35e3a4e3e6558c26bdf10a6f0d7a4"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "02659c4b44e5f40c424a532d4528f1f2140f91945caf68e5d59a16cfce50b99e"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "5352f6ce3122a7d2ee9d8491d35090f006205d4468583f017428f45dc452f3d4"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-de-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "68231f3d7e94c1e0b03559bc67e53011f653b1d8fa9409b9c26c080ea6573688"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-AR", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f2d22dfbcb791e5c9eb08e89a04afaefdeba6b990636330407c6ec5de62ef11c"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-DIFFUSION", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "2c044647af38817b28e8d8822be86b6b41082a90e43d8890ff6b814f781bd0be"} +{"id": "modality-000111c2144b8a1403f1e921faee282879b97a0b133888f0436edb2eaab09c85-ja-BOTH", "group_id": "diffusiondb-author-fce69f90485b9eb97189d66c0191ef5ace2e767c2cdf08975bd82e834e899dde", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "febe897300300b885810fd20b1c2af4fa972eaa2220c2b2caa8d1993640898f6"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "8cd802c177882abc441e3eb8826922e4772dfbf97b6324fcab9bd2cb3f794f73"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ea6636adbcacd968183ea6712cf06aa8454e755874a1a0164254432dbaa30829"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-en-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "acff6c5e4dddae2b319f48016a4db8ce7791686131e11675af4d4497ff0b85ba"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f9db44151ae8ef827982f7279954127852bf952ed1632d5e09f7a238c20c5bba"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "102e6e50f1291d349bb3d167cc718974f09d43d94873e7e54c2523e43aa9e298"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "573bf0b3a6c4fb01efa9dc8e45637e149b9eab6c821bc7fc742ab1b7cdb68326"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "bec9fb1d7bfd352425105751dbfdefebad832292420de0d73d533717af4a57b4"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "90d2eb139de763bb78b26c0dc4a918a90a604ce230b670134f90edf225910365"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-es-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e669d77207a3ed6fc9b3785c0a394c4dd3345abb410cb56f34c2c365bbf0e5b6"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "8457d31362b359d1108530e679fa417a330c9ec2cd592b5c6f8c602b66cabf80"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "8b905382a4514e0da8ceef13168be65577ce6fa60ab5a80e0000546cbd1aadad"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-fr-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "fb3d4d19f922457ef03c678ecf657389e16300bee3e22a9dcb9457a0404bb7e1"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0a4b60b2d478df687f567008fb81d9531575035c51f17c1f7d21a4e9db075e68"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "d057b6715d8578152d0996ad51e4d9919fd8249b855128f35dbcbb45e35e703d"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-de-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f61c37eaa086db0a0f09f459a7ed52dcfffd4f6e4595499701530be52fa9ed60"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-AR", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "532552db945af54f9fba3cc0c6a6a362ced10010eea7b6d784892ff0e7a09622"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-DIFFUSION", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "a2124638ec1b4e029c09f5de479b2287df43838034c2df0b232132de380ba84a"} +{"id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-ja-BOTH", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1be9032fa416d5bafeba91ba821a0b601ee5335cbcae543db79082e768ab865b"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3ff7e35f686ca21e1d7708dd2b489cba091471c0296273e7f1572a375d70fd09"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3ca08cd2c46b911c710fa74e109b1ffe435dde1d3e3e3b6d2d47f624fd5ffac3"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-en-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "fa09457beb9a6bae85550391cd6af5dd294229ec3117dc8e2d21ad1710743757"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b9de5273e01c6cf994e76245b34c5fda4f3ba5264885498b3a239e9ce851c9db"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "4a6d73e2cb0374dd4a019241e67397ca3e5abb90f4eae5768515a0a32fe89b69"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-zh-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f1e39cd7a98a2497c957f78f132db2df3aa3f7fe1898511df96dc2b3a1aea135"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f793a0a23b006da803490b571913b11e138ee39052a0ec69829886d47ea8f1a8"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "17a7c099e5c2cb437eff10acb09c62d681b550871f9256f6a9dccb75311fdc23"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-es-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f4b43a4b352c0c2c305af989ba330e946c6d31195a8a0d183fae346fba4acbcb"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "9a24e6a9c8fe1fcd816350bf6e8595a2731ad2e82d9383aee174b2dc303d1bdc"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0f94dcdf4588727f2616409d3621ab6f1c30b4e69b7198c139a7919e73d06981"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-fr-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "2bb220b15c95132d488e12f497f565dfce20624342d989941f973113e00eedf3"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "cf0e4cbb90a35bcb5ab6c0ca608b6b724c6082f987af44f7d33790e263845ea4"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "2afb5863222a82c4110e26688e6f31af7a7abe96057a57a8ee8f96ac35affc36"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-de-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e66ec0bb5927f8c8e5dfcd5970a491851216c94b0ac08d60c53950c7ad350643"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-AR", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "91f758157a5697c5b2e341d29aa08580d19bc0b6b48b563e52986dbc4d81df29"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-DIFFUSION", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "c0ff4379609ad5b2d9a42af4b0905e3809861dfee7f1342ed031160646604f88"} +{"id": "modality-00021d43bb2850908801eacf99934f133ef37637fab6241b93921ff9dd911a74-ja-BOTH", "group_id": "diffusiondb-author-2d74c509fd3fb870d274a121853ab1b9990e91577cce37582f2eea772692c71c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0b4219df2bb7a712ccf2f7a3af834ee1766c306a50015fcb96b571141bd38d7c"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "94e5102f8d46f967fc3a7a70e9d18d83b73137a7ca367d49c9835ad966be13e4"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "268b316f38e0c1daaabad6c2a20ca097552ca74a4c0004169dd5fe75e8bb0eb4"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-en-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3fc7b6a27c85e9a99eedd4c8b2327790399b291589ea71dd79f642f286d0aa5c"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "751aecbcc1f8be6950916630b6a954d3f54e5e3356a67b7bdae7a0989f62ea4a"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "ffa286ccb6790aa3bd17ab6c6a34b7c9a8ac74156066e6e806770305267a849e"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-zh-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "9472615e18077181cfb3e7e86493547c498f1c895b91c9e6a5fbd569d82e7671"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "8dbdfd489290488a7014fa3e62b4f950efb74d9bccadd8514b8f2631824f3068"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "791fdabe9a04594d8afd2690a9a0d7e7e7e53767802c6b1d137a5161768100c3"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-es-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "29be11c018a54eb9125ec51d7a890cb00826bb2828bda4527dcc31b16c72088e"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7812f0bf463beab6add80fc033fcae49f467dd9a43eda9663c77d49da5f6526b"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "af8704458fe79d7832155d31e735474b4f806e41ca22c23149896dcc95c16e52"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-fr-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "1bc13a57b41819c4ac43ccc69a494364d46f4268901f06b1a1862973b8a89ef1"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "5510c45970f6c70a32e7cb8a411f6ebb94185beedd0edd19654a9f08cbb6f933"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6615c4767be68083c765699d66907e26fb0305e53490271738507f0f876daf07"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-de-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4686ae9fa469f09483ba7b81cad992e0ce0b3bbef636216c365989a49d704d9e"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-AR", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f39fa6e7ca81f8835026475069f8f262d91baea230518aa6968a35e73f5e52fb"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-DIFFUSION", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f25a57e112a6fb70f5c28ba8b6582a41cb05234f2c5cf0797e980f12dffa46a3"} +{"id": "modality-000474caf345c5c5d43f7d0f58c0123f92abd422f9ebe1fc3dfdc5c3386f8cac-ja-BOTH", "group_id": "diffusiondb-author-f77b2e1c41b14e7d4222d7509c8cf7d211673ca52b41f6c5e8399d99cd740e5f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "39808f39fff46dbdad3a33dd90b7636fe667977021c739e0124cdacbf542abff"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "cba5f371d52002177324b7e135ad8e1d9373c3959640d65e95c8d801c0d251fd"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "4e2813a8d807d76f561074bc094bd4393ca70f4492cb171f83299c0f181b1013"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-en-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "014dda320f79fe6fb088560434a1f70ebe639d0c8422e733d8e84ccbc56ddd14"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "69644a5c448c5d233dc0d57140419a6c57cb6536c1626bcea78803698c7325c7"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "15eac0a4d2287d3a9cbb8fe195de290398ecd792973647aa0e0e98feccfec0ef"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-zh-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d4c1bd47c3088c5c18ce42b90e121d3b2d921e5096d1512052659a342b357203"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "eb4e0fc028ad8067f065633272aee00007892f0549259bca4261773887e4e696"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "fb290ea45704ea7a48d84e758128354920ca379ee6f939c8e6eacbd1a52fe15e"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-es-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "67abdb17aaf1132871939a87aacab5897a444b84aac378f74ba906bc9309345f"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7ccfd46036f233981c08733709394e282ce11ce59f680a93c4d9e240cdb8997b"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "03737320d93bfd8e212e5e80fa6581ac8951e286455e88c1ea572f98ee70254c"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-fr-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "eae233a87cbd49d105dcd7ce6fbec31a6f6e221b707b6880c1dd18f8647cc464"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0727bfc8bc2226bc6bd5c7b12a82e20342e7e6813fe48965f068a2512a3ace51"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1a4e4b2f446e0fb08b166b7d8334ed6f29b33e5de0d2c8a3263e0366d6b713b0"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-de-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "d058a65408fe942c1a8be4c679a051edbb719a8016fdbe5bb26049b6d031ccf8"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "33c6994039c305b576ee8726907a85a5bd116e01b375a953296871505c816925"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ad30101fd483ca4ae067e93ce1ad290c867c1114ee06f799c68b82431c723f94"} +{"id": "modality-0004e4618fb943da13ffa80362bd95c4833a5e6c63511dd37f2621a2b9a72fb2-ja-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8825d98397f9d0fa2d5cd9df7ea3e4c56757d0afc1a1e950e1ac10d684f87889"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3ee1f905b2acb868660a85266d9a7ae94e50f9f4eae47ba4f73e7be8026c1f34"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d3962cabaa1ff7f555697181c53c5ba9c7b0826483ca08cccfc23ab3cf112b47"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-en-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "dd85067e3fa287fc8068765289e79ca7d18b42be3f497239e5c350e3ad4e3e21"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "23f59d5fb09f4a595b3deab848c1384619a39f936c5a2cd0557ce735b00f0a91"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0ee2bb74728274e69c6c7e4c15fa0a2bc5b4a8aa05033394ad9fcc151f405ed0"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-zh-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "601a0772265dde17d7e969d1408cb0ef16071b4ee57b8b5443830e9b68398d4b"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ff8f210fc92969aed825e72047e63e622276be8b73844e93f8305dc4e11b6cca"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a338e09d1c652ad7937e13cdedde0665fdb3b999ad41b3b11a83a66ffb8f7cf5"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-es-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "b536fdad308c3e38786314b47f903c964643e23e2ab0acff5d44de6efd424c1e"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "83d225e426cac8400c536138ebb0c0e418e4d7f915b4d5ecbb9358a75220fdd1"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3a5e4158e6dfd6a936bce9d2507c599c3a7011afdd2043f9a6ce7e21354a1163"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-fr-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b97c13bb41dbcc4202cf577b80812b83a645fdd406da7cce4a6d52829b4db092"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "a5d0ccc428b4924653af46dffdf255752c1c7c12a5497fbb62ffa032dce7dc5b"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "b939ee8912fd2b9b18cf3428c77dc889cef86c1733e39937662c357a442cb360"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-de-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "77cc82920fd440eb5e42b785eb6bca4b8e89edba2a67444040d7c0a927a05797"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-AR", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "67f9714ac6b7b41594a3601c277a4652cf6d11831bee27c7e9639a5f62c76833"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-DIFFUSION", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "82f4c9e3b216e7524f408e5912665b95eab07c5054713845001675b8f0080246"} +{"id": "modality-0005853fa3af3861d7051bfadfc8c1fcbeeac8818f0aa438f75dfd3888f834a5-ja-BOTH", "group_id": "diffusiondb-author-4338a4f26990e055ed8b3b0cc70f42046b200fe7bf4b3542d4eb1f55cbab90c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ac38f242b7873ae2c288c7c499a3adce94c19909d9dfa6d6fc568cdf20eb6a58"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "24f8c5329903c2ba76762da1cbb7b27abeb8eaebf221b09697081380aa0ce625"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6c30cc87f1fc55777bf1e5e523b5ff6b1915ea81bbb093e8715b431247e33225"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-en-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "bfc3ee57275ab49ed34b553180e5f691905ce2de28be76865abed1f74eaf5833"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2adee46ca34e168de53104e34dc4875667220a62cf35023903544a98627b2c97"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8f758cdea4a983d34349b2409ddc070804bd405bf5019e7be117cf7018999ffe"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-zh-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c186e4aee7091dfff7aeee9d30bdbd40a55f5ce99ef4ad3a9744b34ecf81f701"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c0bbb3ae4d4c7182d1c88c4966e62d22ac8e60cc54699979c495a93c5f5fb1d0"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "3371647fa14ecdcd172cddbe12a5cbf920c83a8235bdabfa7aa4676f0fd7e172"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-es-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "02b465160738fbd850a2bc1439203831ec036d4851cf38490a8705666b4961d5"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c62e50442940fc629972285a9c7e26bc06acc16c027e2bcd46dc1e34cb7d094e"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "d5e162f89627be3edb402ac3c68af47281e0518f78fae6166f001e5e3365b270"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-fr-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "30acef5c395e371f62cbfd580481e87aeee85999a5eec76c1cc678370257007c"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "3697fea1c97e59eec3d961f42b376eac820e894dab9bb7be95b1bfb1aa2b8c2c"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "eae2bf7fe7dc6e8c574afb755618715d515afb96f9c59b85eb32b1a187e880a8"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-de-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "7f5a02900f12fd2d8ebca3173b8f26f393c98db6ece5fe91bf45a5d23ef02ad9"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-AR", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "7a07a972350a1215d7f0cdc01ef38fcc50f7648d1106304ce9e71ea5a836aae0"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-DIFFUSION", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3696ab4d65bffd14f2aaffe19f9c87126e899cf834de99818533223f17385aaa"} +{"id": "modality-00059076d448682904b4391e48803150f52d53de04683d8d73ec0d4b488ad4fd-ja-BOTH", "group_id": "diffusiondb-author-27747d825470044655d94c81f49a42e0df95378790c6650162918c8a27f18662", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "bb0bbb74e4e2eb3994fe76f257f40b8eb726597719187b0bb5e939f4b0732915"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "93d281abce65380abe0e386c0f8c2533de41df637790060391d2cb6584d18b94"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3e4459f4fea7bba1732b3cf51c88a1bb66020c76a7147917f339bf8fb8ae6bfa"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-en-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "fefcd7c510c8547303d18f58859d2cf8817ab20b6ca8ac96a50fa09e390a1df6"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "105dc6e40efdee5978733d0aa87cbb1bdc22178f99084aab7c2d263ea87ed482"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "39bd225e42d27aef058c17480b2a0c7917ace2a59dfd960f62b76967d91be0f9"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-zh-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "076b352e9c4b2c2cfccae20082b1c8fc8855c1f4d6c18d72bfc7c37b9db4a23e"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "89485665a47410bf2d65b284308b8596c7e250cdbf366a583ce3230aaabdef8b"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "26d30b901d57e5e7c62823215b1ccbc6d81895b210cb0e2738c734d7d6621bdb"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-es-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4723e0e5a113d263872cda7a2129a07c690db504a06a403b74fefb6d89d46d94"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0d5dceabcc5ee5344824a329ad81aaff1c5f8826c721ab8802b43c36ba1e086e"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5acc8c5ad0a280f37d7d4ee3e8462f610da11639647974ddf87440920899ba7a"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-fr-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f0d1680d2dc8f6ca0bbf8420d1271e9a751652cf883b3f217c00eee65996c923"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6c3762f59f9cb562a379af5b657d07b9a8228126e05af41c1993329b95dc918e"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "742d358f153ddb3adb56b8210cae881c3c2ca4d29963fc767a73016ba79ecddc"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-de-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "23a18f8256aa5716eeff507bae3147ac1c83ecd0b52d8d959d36a595016b3a39"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-AR", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "dc2c0c32b653edccce8ef70fca5d50934d5a2769a7ca80bc1e50c04dec6c2fb9"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-DIFFUSION", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6c3cbef223159ae814d8de8449042c3f5cd03f7013a773dc4c84aba82658dc6c"} +{"id": "modality-0005f2d212bcc6f0b5459a05e3edd6614f1e4cc358063dfc10993a4d89713a31-ja-BOTH", "group_id": "diffusiondb-author-424be1751025ea07ef1c95f962b6a2204f1d623ba2578080c40708f39401bbd6", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9c33802e725a15de4353e7974e97ba579cf9303e4ed4c9d7d4ebfd2f49dd69a3"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "43ed073f2d78bd19dea129d719cd846f695c9365aeac46d6c763887ce10cbb2c"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3c5782a1754cf2110d4efdb5871d347a6239fe630b765cd1a4e4ac3b0628d973"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-en-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "bc0c68f86a2246fc00f6fd5a1f0d70019b8e6bb1d404f6523ededf28d7c6054b"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "689d85a2a9dd6aa804c42b9ab261544fa7c305f3ecb2d410668f6669b9508327"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "4fae2e034263ca1c7dc9778e038598c7d93402791f78c2e506d8d7170d698b35"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-zh-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "28a4485c048fc84ff54e2c9172535cf538bb7f2a230b6eb5c7088ae81980dd5b"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ec0973cf6a46468e968933bc45a96cbd495377a88e0d9c4d5639b2a1b8f93516"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1f289e87164543cc003ccc55ab285d80722845fa7914db75627e351192c1952c"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-es-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4317104d3446d0d59c9d9a635f4ff3715f50dbffa42a985cf5fb33b00a350b1e"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4e282f3014f2d8e61c291dccb0e883ba63136230ed186a86fb28b438832bbb10"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a61168b5175cbed9a9d1b31063fd80856fcbbca8c0dc5ad2d1eaf38d51f3299c"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-fr-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "debfaa25d05a533dc9a76ade3ad86db376af3d965967e46d5ff997000ceb19ce"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8adc1cca0b37b39ed1500c5487fe8116a1094b1caf56647dca05ed5a79a6ba5e"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "ef2400aadcdfd257d76074f831ea4be54b1d34897e53d8d2642bb9e0d59bb005"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-de-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6ffb01728e12790f7e4f36ece8c43f3e4df7d42a6b90db396c824f96261cd0b9"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-AR", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1987760a4734fee0887ae1fe3e83efa0bbbdf8955adf05842ca45da9b9e7c3a5"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-DIFFUSION", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f291f6ee809f590f60bc3c4a759ae43409da2f54e0c8d778d0ef37d9eec98d03"} +{"id": "modality-000685e7b44be48efafb32ce9c967db246ca298a837fc8ba2f4ee662ef849063-ja-BOTH", "group_id": "diffusiondb-author-46618b47f9f419dab488d8e893abb4aa5a871dadab0857b3688255c5e51b8afc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "745bef70dcbcc3b0c9e90b24c09ce804caa8c84ba58fc98ca42484dda456b268"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "830e0a71b9fd311e6ccadc302391762a084f3a59d2509572f1b7cee1b0db3f03"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "db1e23ae0333364063e30a417d6c51492c4998b8ac20ec49e8197efb15f95fa0"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-en-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "7e1f8dd9a5a811fd871255a59a0b75590305e8cc1d3888374d9ed9acb8338e29"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f784b29d3d559d6911c9de00253053fffdeb48e4cf45a1aea46342a58db9e1c8"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c610744ec8c76313273a39684231201009e0042f8f0755da4efeaa052234910a"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-zh-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2dab77f184077c1ce59be7bf6a3096169df7effd2a0016ec27e454cdd1c5b7f3"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c3d076027b9da28d93d4a06ce16ca3f7ead7b65dab1f8e914fdab3b82807f17d"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "5e79fc0d40010a3b547fd47b582248ef2cede729f8dd1aea8d02ae680e036d14"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-es-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f5338cd3431c0e913dd3ec991db8852d48699b48c26ec1a637c4af32aeb4a13a"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "70006ee1c13f33943b67cc9991ed9d7ce6d44188c15140218906f0e0851e6308"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f26db320393e419c497b4847f9e549da0c0855079249dc4729cb62320e17e8e4"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-fr-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b4f8fa6b7f5c0806349d67efd04d4a646a51d078e11f45920f42b6a913363a44"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "41ff75006a0f1a32756a8b887dd3f146a4a7b58b42b1ca0a675ba80387cb258b"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "dcf84504a08580376aa08bd990f7c63f95d2aaeb0014ab98d7770747eacda3df"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-de-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f282de65f2d5a9161315cf5f4694b89f071125d543bda7ad75bb32eced604d14"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-AR", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "540e4035f816ef41378d248e0e98b93befca370fefb0ea801d9e8d72b13fbb22"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-DIFFUSION", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "24328be54e95e5882b9de0fc85ed85b7055faec21c0ce6839855800aae186e85"} +{"id": "modality-0006af4a5f5a17fb8e642aee5ff8bd981b764eb2047412b6b51ed445a0531435-ja-BOTH", "group_id": "diffusiondb-author-178a3b241a97bb72b5ae32c9751b67b4fd3f6e0e67edfb35d663434107ffa892", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "fbf82d3fd61613c9615070cecd9e5313a8e6a33d703d9ad5acaf7df4705259a1"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "fb5443964bf5da4e9efe1f25af357d8a247620d8a7bce20c860de8be24db0d8e"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "985dd6af0dcec7b6bdc2dba684b4617f274bcb5bf79f9ecb54abaeef86dd5b5d"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "38d8befbed8f79836b55be93b85696909425f72b5f7d807587b4db86c7e95873"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "e2122a1fb273018a7d0df0284fbcfe50e9a098620a2c91d7e76b8feec83892eb"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "743a12f348c1fa18846a38b7f610cffd13153d9e7926e8e9b6116109d538941d"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-zh-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "4fae0afc84b8c5b90c9d52b1a7fa06805038b3b2d1e29cb8be3a8be71896750e"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "659cbe3374749ddaceca967a0e6ed133e4b73930a272ca93d1f794ce5d9da891"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "3daf0010b6d7987ca59fa00a2b4835479124276fe172678522a2153ec4c03750"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-es-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "40df24df25d4de7763ac746c0602fff507302e19443598cecbc85bbbd12a6a66"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5839b2307544ebdd5dbae66cdd7e6f923c843c0f1788c1f9da59541c6587adcd"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b79dc5e0b45db3d7cdf423d8fb7af02e3cc1f3176f35d549f7f4030f68f129c5"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-fr-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "9d4258f59811a20f9da7051210361781fd35d3296a30afbf72a1a2f9c2a09e08"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "d6d0d17935491aef1bdf2348e742c95dd6695031e6de319692151fc08d01bb62"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "c82501c842ee6fd6138048c2d80cdd8ce2c6ec5451cfd192c6f5d630d2026181"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-de-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e2d4076dde6dc1576febfe28c88d1fe80e6e486473650a66ef565e39191937fd"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-AR", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "14af68ae9564301396de4766d868627ddfb7ab263485aa5c31dce0f3f2675ef8"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-DIFFUSION", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "e3b048ac0666b105f09d05aed9824c19081f3e726403322a50e6d200a573424f"} +{"id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-ja-BOTH", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0f25006627b11159a64c85c1779f7937062c00cc17116026f9d503cf69b2eab6"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "a1cd494de34b1f25e0371eac5000fb3a7d73d7ac9b54fa9e4f28287da90b0622"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "82483aaf849b7001bad8592b944c62ba92cb682e4b72f96b712cc7893d23e7d0"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-en-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "16b35731031786e9bbb7ec479ae54857f9d31cb6086f46dd4bb4944ce2b50209"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1a74861835655e897353ede9e0c46a544702d263d927038d5205b65983b50c32"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0971ad76024bf1879c0fed7e2b0274166a6f0431152ad04aee191be83449047f"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-zh-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "509025e539fed2cb3feb4c74fc82ceab90f8e7e10726f4ba5aec136baec561a9"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e706caea2dddbd83d49e9503a1b55f9ec02f6b7b08d057f93aad32911b3b1737"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f3ef9b7377b2c7493f9e68925430ab5526c9f8b49ece409a20f29173135ebd91"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-es-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "25fb4a1e0c91db1ee0107e60c75b72448ffb3452115a43bc1a8a37dd3068f5d7"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ac59b80e3055ab438d9090ed44cb8e86c14992ffe3f33fb676b41e4658861922"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c8a0cf91bcdc1c4b56b1f0bc7b48c3ffb8cbd04e55a8a04be7818dd0414c7dbb"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-fr-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e357ec0be9a76eded6ef90d8316821a67b4ef77d6743c2e5f4d462f362cd2e14"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "24356f0021b65e2dfc77e3c469741965ce07d874bec6d0e34e80e1f7ff17c559"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "47a91228811a29b02b486ede774c25e2ce38b798d72947df98871471777094ed"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-de-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "2781229d121028ef61ac328df759275ebec23c43436f0c42e48ae215d8e169f8"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-AR", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9a5986a9fe7497849ca27257a67c434b386dc840ac3b5a9ba48504081d87b3ad"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-DIFFUSION", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3028d6bc9fa7c5699a34b40e35babd57c9e0380959cd88610e3ea5118f21c656"} +{"id": "modality-00072f64ab281d390eb788c7d670e28f8020ee0e9fc6f03a460236bcf8582cb5-ja-BOTH", "group_id": "diffusiondb-author-9597617ff6925cf9aa8f840272a546e2c830d1649f0eaf4c52c1a4516fdb125e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "2f08ca0955694426c0e558d44a99fe23ca6e45d353e93231d04f7206429c7a24"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "85cccc0e8cbcf4a4ced799a850453c956f9818318da276e2848080f5e947662a"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e2e5a3304a0af148378e3f98cf6a7abafdcae5529a9e970aa02bd1ba0871b3e4"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-en-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "8fce5bf652019f2773865ffcb44912dce4ba9d5c93b3632550ab5f819f11d674"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0b21448d52cf819abcb56bd06406f480ecaccf927b0aaa030f6bc657300d8a20"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2ecefa79dda22f58901c1365d02810823c6b59d1f02f567150def67920afae4e"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-zh-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "9ca1800bfb93bda39b8af7cd057f54ccb3808d4167a723b764487fa18619c132"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "3a8e7ae001369538b049e97b3d51ef98007c04b26772849a212753264fc53b65"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "91d201bc8c6ac468e5815de3df65e5664375deac9625e852abdf12c821962035"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-es-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "16c340a6926f34b8c51919ce92d5245fd66d0ad8afc9517936d764e12bd75b64"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "1de3cca2cf71332703a5a30d99f814a97226066065cf987004ed2df8fab24906"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "fa58e37a900928e58eeadc314b0743fe248e8419fef73101f53fe74d9955dd29"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-fr-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6c1fba8a00865003a1e9890934116043d54461891e209847e9545378e5f13d26"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "677cee705f7e2379dcb54a017a53c25b4638e54f1ca5ab1b38a04cd53d4d302d"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "3a7aea83d73435e53a5808d94e66eaa6cf48172cd7b53170fa5622bb0207a9a0"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-de-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "eb4be110a2cd3d92e9384886e976fc9c46a11c7f066ee6c6531b0ef33bb5bed7"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-AR", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b25dec4e0ba591837326259c35486f77225cfb10049fe07caaa5efd4088aa3d5"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-DIFFUSION", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "5d9f9615da349d35a42543279c75b424ae833e971d2a1ac8c73e827d999dd868"} +{"id": "modality-00075e3cd21c4aa08e1a508c1b003c2d2a9dc507c999c9ed0b8176d4ee547228-ja-BOTH", "group_id": "diffusiondb-author-3f6eb45bbb60a95dc0e215f77462ee5de2b13f04acfaf49bbb7e46738deb0996", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "67e0840f4d64220b6f1cc2c0715417e831c397da10559960e4ddb868abaf3fdc"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "877d3f5b651dd70a4635c7982eacfded244ee919c09932a33862128314968104"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "9dc2ef3b29e7bd80b0c9bff507d39cfb6d73ed4a589fdd83a47977bd42e9b682"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-en-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b2df6ad169594833a9285ea89ed67db0f9c9fc5d55e02c71c3e1111ac25d98e0"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f12996b5e12cfba4f9ee3980d1e6ef2de9fdc5df7c60a4d29ecdcbbde324f55f"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "52c5575439d605b3891f8a2831fbf1b5d1baf878b82206f8840e3b0896aba489"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-zh-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "6d63d4d6bbdea91ae5c6bd39775268b9c51b8a69869c8dfd978f992c8e535a6f"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "b6b740432e563b840a5e96855c28b5dfaa142179b07f6dfd83758bf64d890533"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "68520c02e2e50fbf6f63bb8292c72586253a857f4dd1868af1eec9f1605d3f17"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-es-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9404e908e67f5569442e4c37b90a416556de840bef92f416797ef034ab6b19ef"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c07a1bd65ebfffea8fbcc780c47c0527b60226f49dcce639a2cb471e7a052de8"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a3fbd73afc5c3545aa1e3b04f0dc830be5133e6d9b38de99157d5d5ec069f7c4"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-fr-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "461ba7e64f302916157961fb83d7b4b6c241b300c2202ab35e30c567cf7f6ab9"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1231cdc364bdd69e5ad20b27860852a1852bf9b8d12243e5a49e439023a832a0"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "110e02c6a8967862b489a33db9628ee278b69d75b0d1176cd208f285cd810485"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-de-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "3d20ca443a1837f56863dc4ecfdded5681de25c094d0d6c2fb3e5c1f554d670c"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-AR", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "389897c2f8b1736516e7ec71e2497f670e8af695f87dd4f9935300afdb54f1c0"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-DIFFUSION", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3fa97053b6207425ee43914876255fef5f4890cd89810b142d2d9db41d1d1f4d"} +{"id": "modality-0007a3deb4062c7dcf296915dfbaa9b829d6ba5f5ced987cf8a280c3eb9b3b44-ja-BOTH", "group_id": "diffusiondb-author-fbfbe11e6a262738ed99fedcb7749039695dcf342b178877ce580fa03c0c9403", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3f2e30c87ac6a088a93f79b7aa4356fe539245b103c43598a22245fc6e500905"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "5f582f14d957d57201af81a0c3ceb9a0b3727a77699b4737755bf8fcedbded0b"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d65dd32e4319db8b943abf6d8fa14e7022bc62d647503478d708d736228d795a"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-en-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "de99c5f0bdfc3406e61b6ccc1edff1ee06ace47caba63d4dfd7403c074f599cb"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1f644fa8097b78170fffa93f6747e144adb7ad2abd32e42b355dcacb90f1f011"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "740bd11b15706a11930d7384cf7a5f4aa0a4a9c1cc8347ab787c06abe5bdd618"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-zh-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "146585e4d11805088d2ff7c8bdc8baeee474716dc7cbf6586b76b3170f4486d4"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "962ebf0325de48f9350136d65fef2f16dd828c0c1ae9a4c4d72bda67eed1bf00"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "436f06c97ecac6c3bec9023c69ffd32d05cfc02328ed7d9d7b0d6710c34d362f"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-es-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "88fd5e5555e1688c9ac0400221f4c1393eca471d18c7259b90ccf87d9f479a60"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "59352988ffb9345409168bc78a301dfca987c26d463bcc73b69e32c7c4354eed"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7cf41d59f1b18350556df1cb352e95700716fd5df86d6ebe6760ad733c501483"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-fr-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "43117d6785c7d24f18d3a3289f3f4b83700f77ae1b23b59b2dfd72f3c762f4e6"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e6a0f94c2fe4fdf5bf8121fb5cee44b64c95a83d40dc8cab4f429dab1e9365fb"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "899ba7f3a15a3a39b1c3915f28879e4372c8d67a7197a2814e1e6477f37ebe69"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-de-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "26e669546e9168222bcbf277057fd9fafbb309dfd2073df0129356f85cba98d6"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-AR", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "e003f588eda05cdcbf41ed152cdd581374e81c69ed0700f7597e85dc8128c509"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-DIFFUSION", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f9f8c6ca101441e8a4071feb3b69549caba19b0b2a264c2e3f41078efd47d827"} +{"id": "modality-0007da3632c3baf7e40ccf6d01cf7dc3824595c86d7f34d5acd145ccc261b180-ja-BOTH", "group_id": "diffusiondb-author-89ad2dff4763e8f1fcb96d015af10703eba33b4632c7d6b9dd48b68ca6fb1b08", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4f984a6c74df4b28c920bc68804e6b156fd4233146240861b8c71b6bc30349ff"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "94948075de6da4256dc0428c63dfe5165ae7a62182f5073c3c1a5ff7fd8e1106"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "cf941b7751cf2488fb675fe77a427588b453b36107b9c0e3cf4024bab34a9fb5"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-en-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "303f615954e91e06ab33f8291dabfca4a6a2feac03f2328dd2e3715f1566fde8"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "12f524b8f70098f6e09ee5f4553a6fabf4568e6540a56a8cda17c9dcb0cddfce"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8159532bdcab1597bc67c46532f9e08814012320219b7d3aaf875f74d9d912d3"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-zh-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "aaefcec2c2c396508090dc2c76181ad6bb110c27b5e29bf64a346740f40ab64e"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1d5ea4e174dc6e1737c60a8787a395d8350d41c8c333abb00116f825d974cd73"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "48ef7c80374b89c7595100e3474ce63e09c8e233c2f3681d1275c67d54c3981d"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-es-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "948cde26ecb40dbd949338b02558fce5c3f98b271f95c51d10a10d7a95c7fcab"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "768d2b3159e38e42f476d55c42736c4874f8ac9247bc29a39c47f327c7004b70"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "37122e5c565e7fc7d57155968df374002880b6aa5ce0c67850697e6b484f88c2"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-fr-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7103ae3576ea78bb88d6df32b7ff4ae4f8617b417a5189e616820cea1ae20e2e"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4b63c4f10f9a2d4f4403ce9a74405f21a40697b766b8b7822832727dcfc66a4d"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "29120c384018f79ac8e2ef1b32680a78411befa0e677540efac73614ab6ff9e3"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-de-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "26879317a984045c02d68f96b2da4a77a950587199bc24147546536676f03a91"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-AR", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "430a628c2fb2e2fcc1de97358362f50e90a9b12ac52151da64c952039201b176"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-DIFFUSION", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "71c5178afe3ad6c5edd79832e7f137522ee39c8b40a25cde43f96fa6d425d718"} +{"id": "modality-000837e4c3106d47b1cca4dc2f2fd1ae2a879b8e49753c21594596a07e9ea142-ja-BOTH", "group_id": "diffusiondb-author-8f6ef7c710878f710a4572d6d2ba86888403d7f1f740162ac824c6f05952c020", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1675c130151cca52634b75b29746a7a205cfe7ddae85e679df9f7f955f88ae43"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "c606fecbeb417d8f8aa89bfc2cfdac3d6f31e53f99374ac20ab48dddb6ebbd45"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "84d8b85ed102426cda003f526d7997980286d8c9aacc3eaf5b9f223007997e10"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-en-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "780c46a20021de9c51a1525734a20fa4c5fd85c212c0919df260ce3085a33e33"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "05042d7e038745848cbf8bec35b1def0926cba8a899f5594cb13776b8241f0e9"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b39f4b7314116ef5b92592ebbbb5090cc17a46d89118d36f5aedccbef6fe117b"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-zh-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "28fb656f53aebbd04e698c6edcb95827939c6df89747d2848c67e49ebad9b9f9"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d0c1c938bc447121090ab900ea552a56ec3b24d552602dd1839644bc8777bed0"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6ea6862b35dce1ca7007022590e385248d410385a250a1707e1e7abceb91c8a5"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-es-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c7ca0583af45f868d11f8f10b634db66303a4d705b79408ec688b0e27f945211"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "21eb2e603f35f19c08068bceef82d62b511efa32b3522e4eb1b2d47f61eae952"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3c68fb2e4b76a808d5ca4d9fc99bc50ee1dbebc0fae9695cd3aea0c64333ad90"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-fr-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "1029cc4dc737a11d61499cd12607aceefcd575cd6a416f5e723a8e162a6e758f"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "ff6f86ee5e0db909e92eeb79e300c75688c8a8a29a7f01844545fdd06b5b39f4"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "423bf0dca93971af1aa0d702c63b7dbc816ec2af51da1e0a51123dee6bcfa55c"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-de-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "2cfeff14c9b46639e87b581717a19dc4eb906b21f905c8e9a12c9f9b10cca0c4"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-AR", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "78ebe8e77c8a6cdcd916dd575aac1a3ca5a1054de59cae764506e94a1f9f4e6f"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-DIFFUSION", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8f6ccf70c3708e43cabc36a47ce4cbc577a5bc2b188182dd549efccb046ead31"} +{"id": "modality-00084775d0d7926b37583c48818903625828e65bd27335d2e60b743574d8c81d-ja-BOTH", "group_id": "diffusiondb-author-4c8c425247afb97404efd7736f7850098a918257fd8356a00806c852fb5f44d7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3f549dcb357a9764b1cdd8a66b5f529be2ca95b457dda165a6c7ed030d28c3ab"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "aa7e6f9baa7f97cb16b76d57f65bacbf42a381232107da33fe0e8400b408a0c6"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "96727054d8f3bff34cc43aeab39c09cd91c20d146602a2d8da6d028b4d1f1eb2"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6778af3d2e543e2f1c90f5954087c7dd08d43861f773d058f80e81dc76978c23"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "afdd7a8fec52afa93d0360232ada0746765bddf69e329e47e77c1f369c8a7d57"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c10b6d71fc372114001019d84b74c3adad5f9186297d4007f43720b746518954"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-zh-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "cbaa756c4319b18a1ffcbfd5c809086916082cf40f17d747bfa5118e0ecab432"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1374b3bed739a45dc08b744d26142bbef34ce2e9cd362dfae35e0eafd768c10c"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "be5a1abd6ea2313f742816ef7dd50fa7f8682fb152de8ac43f9b6dcb14d702ea"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-es-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "51c1763388b196b692e9571214b6e00e20162fe2a12d105f70dc2a5cbc4c7edc"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "bc0bde7964095cec077c398e09c7e424dd86e22a4399c68ffb1262f01d821b77"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "99b90df88a09eb320f4ef546f6e76364ddff044795d46ffaa2faa3bf4c7a48e2"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-fr-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "dbbf4786ea05502ceef6ded78921ac9270457cb72b1e6100cf510e4108385f19"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "209b99532d0acc55d07c5ba35ce65e7123e4f100ea42161fe69c69f46b0dfc23"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "c6976c833994207343df564e4fddf11d3d9ca55a6a1d09eeb7e44ff1db05b83c"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-de-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0d1ccd709ff996440c5567e502e618af51f7384005a0fad8ac92dbc80beb679b"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-AR", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "937a8f569bf50c1dd623a6ffc8c1a4445e8703134463910fb3337246aeffec7f"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-DIFFUSION", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1d6e09d8b8c3b2b3f377bba86be61587c228d71facea0bade62f1d3ac4306a3d"} +{"id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-ja-BOTH", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "56921275d2b35e05023ef45a831670fd32d8e78f0ca694f34f6c019ddc4b357d"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3dd627ff2711ef3b973e56acfe05fe97c9f63eed86420d5ed46f2c03ff705e94"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e58a623c94b4b1b35737b82f0f8e4ff4a2cdf7dd970bc7d857dbd86d9f807ca1"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-en-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ee694476fc781153372a8a9267e70887f63086c6a5f522ee0cbea0f30428df7e"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b6bd87d1af68b869c81e7f3fa5cc32df85259b96c9a16af58634dd4a6380fd88"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1ce9ae3e3b679ba1bf5490a4d24c21badfb017e053902dc8cb16122a4c5ab7ec"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-zh-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f60373c61a78e6a4721ab1c8b54c108b44f3b823465b40664999be427ada9cbf"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "314e9ba02d3407a9e30d998bf45255ef2d8d1b677e622e56a01590c38e9518e7"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ae77f38e82e75f50a0438a95e12379f3591235d54c37811d7bc778975b8d6c2c"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-es-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d863c15fd0d44389cfbc3313b18e720b6502d5a66c238ea900f1d1fafbc42740"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "721dbdf6ace4222fa7d5b84ea9e179aaa719e03a684cc328869d38a2477d2ee3"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "90b628d984f12adc3ce08cd36660fd298b6ce8fd9f9395c428df866b4c6a1e58"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-fr-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6c1841d76c92f0b61bb34ce878eaaceaa74f625fbc182f7c893f02f217c44375"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1a0d62a11a1958703c7982a2663eaf473b40fa647b81dd5c296aa5b948500bc5"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9f26045bb0ef0b32bb9c28f056e89aa0421ca68a00ff452a17d7d32894a119ff"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-de-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8342766a840e06f0f67c4e8d2ff0ed92b2184c497e3868013c790a0223b983e0"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-AR", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b5e2c53ca02af65269223883d9a43369e33a2ace4c23ddc912ed90c8119142fe"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-DIFFUSION", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "70f96613199de8b6c938f91db93187f1bf40caa125aec437afad3de5ee65d86f"} +{"id": "modality-000887ef4051f3865d9c6021e565a76d5c501fb05bbb6c2e4a750321292c142f-ja-BOTH", "group_id": "diffusiondb-author-0ce8d5be188f13668e091f4557ddadc16b7b31ddf467b130b439d7b07bbf9011", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "19d584bcf855cd510f287c7cf597c8edf823e1a6eddbd9172f50f80fc1ce42a6"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "8e0b12567d4e450a78df57f806bc5c4b3bb6a457a09b9cf2d52da26bf830d9e9"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e023fb0497af6b72945ff3eb2a245e1fa960b25aaca11423a8bfce72dc416694"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-en-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6518a03868a420af6c951864500b8e1dd50f58bcbc04fff02a4954fb51577be4"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "3e49d8ce52b2e5e15023516e4e38f2f10077662b979f642d45aa70d4d65aa0fb"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "ea2035ed1bcd7a666a255a1855bba3e57987eec1b3fef575f4b75546800b83d2"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-zh-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d86a04d80a91718308ebad134ee6242e7fe103eebde189433e10be3f69de0520"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4c46a0ec7371c0d5bd3fa9dbb3a14d26a3d7e389bf388b28d71fc8e8ead4c070"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "2e24d289db6b6c3979be559c7de2566b807138652b4a549213a0a24e0ab1df34"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-es-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a9c62fc967a57030d82c7f8d4cf17d53d495aaddc73d6f987540d71d7c8755d1"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "8e3e47986085bb398af4a760263f0c320ba5d289ab2445af2b745dde8e1b6b57"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ab8e396954c9a443230361e740792b57587541428a3efe7fa822c5e49d134bf1"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-fr-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "07176020ab80a1f7e46e9e646fa862e2d31dab0f080e0baf64d36ce5dd4529b6"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8f12d0b1dda4f5bc8edf62f6dc2bbed6e1f57ab04ca29fd78c59b3be27f3175c"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "48e3fcef312b3abbeb8a652b5263e582e756226643877a9d56bb7fa623178f07"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-de-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "a283f943f4cf476ebb980b2899ed33c139488ae8dff33701d40c0fbbfd0bf4e1"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-AR", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1bde00d90133b5d3f5bbaf4372b4275469391ae0709c6a48728aec9817ce7712"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-DIFFUSION", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d75511e6106d0e2c7ee97727cbc2e1b8b6ca0814ac3c352d8e420e1b4a510714"} +{"id": "modality-0008978fee2aeeec77f370c5e6533e2bce7501cd9e453ddf9956fc8b9b6eb041-ja-BOTH", "group_id": "diffusiondb-author-c26407723294c74359b1285a1004f149b9d8834daab82c1c7d3742eee55d749a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4a13af2bc3bf26465e04cfacb2dbe0adb65bdc3188d7b4bd1a379ce39cada695"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "9b9e94af50f6e9265602bcfe31ee70f9207b9a47f72d5aedf3f184055c5e62a2"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6242e639debb58770935b9421e87fe5c416f29d52d66f09fa0a90ae44d5cbc7e"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-en-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b45528688353df1f913f9c341a9a3e6e0f735f8f5a9b67d8ea19b1c94d95df76"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d74ba2f305e5ebd7abba0ae82276029258e2d39aab37e5095d484797c92fa8f5"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "15237d4f04756dfe6010ba7a035408b933d2715daadefdff0077660dc6f43d7f"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-zh-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "237ca12efaa3cd2e89e7156e41c00b7876e6f74abfb0836514060d1b557a28fe"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c273bba5a98573c831ad6bf41d7ed9e3cefcf56b456f8e2483c3718bf46933b1"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "71b73da7d2cf60fd6203e1f35cf870a7277e2e45085629dc47af569d3eb94088"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-es-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "7f3346729a7d01ae6cd62ca16b8c7f0f6a6ea93230bbc3ff7bca00d618b9e876"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "d09441dcda4979b36d5491005daa9b94a98ded7961ed0744e27b7ebeb681eb0f"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a3ef594e12d05a9428d572c6a6cc12a361905ad7a313b0f30595be919ba126df"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-fr-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c35f1efd934b2003f78cac61585f51974091f98d7c806c8f191857ef064083d9"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "b9c75f77c8492b4c59daf733f49c9a24942ea5fdbf1614f87eb877d75353a84f"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "c482caa1cef9e7b64f1a1bfec32f98a0c6187ec8395d39fca907be3ecde280d5"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-de-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0c3dcc8ee6599dcc94e79b33156ffc211374a72923ef9fcd1e41792ecd13462b"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-AR", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "82b8d3d2e8fcd1202a45b2ec7f66a3ec2a66b0ec094dd4606c82fabac723e583"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-DIFFUSION", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "31153bda8ff0c47c7efd3aea1da10d7de81b0bf4610e24b8629c448d664ff805"} +{"id": "modality-0008c8f2070ee461ee3f3c611f031ee32a4f667fb2a9702b6e2ae7030dde9339-ja-BOTH", "group_id": "diffusiondb-author-8cac0f87c5e9724c509cb26bb3a8c443460aa79e012cdbd3f6ae1d263305cdbd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "fc160fb5286c45f91d92d9ac23a9366463cbbd12cc5fb6c7aeac5b0ff16d5ffe"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "357d547464c01ddd540989acb78c13cfb702befc196237205544f7dd749b2afe"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "29b22cb1aeccc49cab7fae9d34dff57e363c046fac57125c73ccf87d213bc53c"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-en-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d337de2127bd5647f90200e07c0c15713fb91737e036ce77b4e075898b63f8f6"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "52f418e8390a9c069cf370e1e72c44cae7751f11dfe2acfecd652b2859fccee5"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "94435e9f549c8d7424da68396414c69a066a8627e9b9ca5d5ddd2cd6564965ec"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-zh-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0abf30baa2d1a613f757f957354c12ed460ff8105982bb8f646479c3b6b0877e"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "309489bb0ec91f0e30ef44406aae916c30433d6ca7dbe5a973e1a6658549ae3c"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "0658d25fafbbdcf483e65d5a18e2ea0b29e69bc5f801b366e37b34a5a6ba70c1"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-es-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "210b5099f48aa30adea3e4b98c09f8f831bd607b5552181244b8721fc9cc17d1"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "afee15a243eb8e9504aca9c10713fb270acf0366e4ca907de9852934378d5b6d"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "661689fa687c4d1a3b8d8448d14186e2fb33c5f9d4e4a2f477ca8a682277b16d"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-fr-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "75ad3547addc46f22becb7d2e6cfb96a63aa8a35952d1ca6526d3b942bb4c632"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f530b09c2baaa6ef269f5f264460f7b61d4f439a6b88a1823dc533eb3f2806e3"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "127c8e5a87dd0b92d8a65d182ca2ea96a56ae3fe14ae57ae24f71f5dfb67c073"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-de-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "687587b95ca156ab7585b587ad482c5e7f147e8cbd5080ebd98e21c8d256770d"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-AR", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0682987072c543bf34339dae1598cacf19d30f5c7e3bd503ba6a034823e8dcae"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-DIFFUSION", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "a33ad56329f277dee67dc441e07c5351f29c4407b31779284ebd48cdfecf23a2"} +{"id": "modality-0008cdea1bbd91fdb69b29a8431e6bd93a5756421dec4874f6026e5148b01184-ja-BOTH", "group_id": "diffusiondb-author-48b9824a0a8110fbbc3324aac00f0f965dc2e7a27cd1b759533c36a91d0b7088", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9fd0c1aecdc519595a1e5d0e8cb1a988881ee345515abdb135667df455ae5922"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "426e63767b23d351a076cc05875ffe0a41c93f1a86011b1cc6040648fc90fd91"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "49a1048a533d23eabc2ee2837ac855a7cb718fb45c6df626effe63d8df9d35d0"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-en-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d621ad059a8fd2003b90d0d93e3649fcaadc8b33f5d0c5790331d5f8f26f0636"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d3864402dd75945d012f988d8ead3afbc3ed18614733e27cbae10e049f9eef60"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8822d1377a02fc79065b0ea441bd3ff7621b593245299b0b79f1af7435045678"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-zh-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2f9d934b1c6e38e525f5bdef83466e2573a0629e798b0bb45b7198e5b8c3cd11"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "3f9efa970ebf6c8c11454ee3221248c90e434bf7ad5cf4c21f8dfc6afa4438da"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d97d42d20c8d5a94f81e444fda5572098ef8c6799a68cfe0f83b0c2d7c8f53f3"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-es-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "2dd87d1883b7dc076dc71300fd283b901fe60ce35f8ecd000e0cb466c0af7ca4"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0aa262ed962aa4243206ea8696524426e9124d7f172740cab9e9494e5f335370"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7c29c80a5622dafd7195b7354aa174d00ad227b75f160e8025ed0aa5b437ab04"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-fr-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "2ba815283ef50bb972e035d481fda35a6275cc0ac1c4563d5ddceadcc30f5968"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "2ce87e356a7b24bf9b4003e4d814e6792eba737e35a719f4d6c4f187e89b71a7"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "49c54875174a36ba89a990ba7ae0c8d200a9b1c060f195a4f96dd9a4a4a9b542"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-de-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f2764cf26e37a638231131988941e07287af725e22faec90f28d8316ec745414"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-AR", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3a4e2fc45bbd2797b30183719c4f09705bee70f36c76db83d0dfef86cac41604"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-DIFFUSION", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "451179387c5d068956487073e44ebab667b1df830256cdd4cc58ec6845ca78da"} +{"id": "modality-00090a8776b6c538a67fe3ecfb6594b3ad49b8ac5a65aac1b7183b315d03fa97-ja-BOTH", "group_id": "diffusiondb-author-c267953dc59b2d5b08145b43ab4de6d1dd3db59827a3acd008209d0386b285f3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b6f73d644eff7e78736df5e372b922404c903b12da6509273d35b4e9a682c7f4"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "391a28270aa9d53c47071e102f9033383033bf83382e2343a2b73f3b1c2f336d"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b3ac35ee366f6b533d6ba234ef52919cb78223712e5fb7cc9d6b7f6470cb7147"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-en-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "7091ea4428a1d75bb8759fd524e914c9250dda2160084a0a9b29df424e123e16"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "a65386a9dc4f7937231bcc1292c3980bd8107452383fd374bfb79092e39f26f4"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f248960d2475715fa227a7bead12e6fa658bda953f05c3528400cf77157332ac"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-zh-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d18804a02d62aba4e2a06af6d3ab9adb63c0d8bd750dc301d716cd68f86594ba"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "731fa7247a336b697a3c5e4b1d09401e34efaca984d16187bd1527bb771933c0"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "368f160a31b35af870c7d9cdbb57b57aa53d99a00690148646f9e7d4121d0616"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-es-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "107ca806810b341b4f2efafd0d583aee5b974caa0a7e21ba02802c9e82984c70"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7c1c81268778430bd485e6d73019e47efbde05e40b62941acd3ba322137e40f0"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "d189394ac5607803a5a72a980ac8dbf3b9bacf60d7c3307a7433ddd7c7dd286a"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-fr-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b99a19ce2bc935eabf0343632541886152baa746021807692f9261842377620c"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "822823dbf413f323d5e74fa7469e5f93a67ba4bce0b266a26b7c385b5395e7db"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "b057fe21f2f78b51519de0a70a9181470f2f276ef21fdcd8a2737b05e7c3facd"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-de-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "25178c171ee5293bc1ee990cfdec526ee0533b8fe075d612ba83edbf3c2fe78d"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-AR", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f047a47c5959363484e1dcff256ab83d47cc17e18b27214572198b53c333f96e"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-DIFFUSION", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "dfc0f88fe6cc2d26cbd6906ed782ea43d8c29598ff36bee8ee32528329d27747"} +{"id": "modality-0009654489f062992b3054fa5a2190770e98494933fcc6268f24475493255e3f-ja-BOTH", "group_id": "diffusiondb-author-c9e6206146057a7480ff0adc3de7735057ab020bddff181215ff07651fcb02e0", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0b12a442d0ca781e3125f05afe54f0a54bd52d68c7a43b511d2a5343eca502ee"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ce9552e24282a9f5871cff35fc0f9182fb246f3b7831ade9303a91d0c67d6b88"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "f5a1b2d1d0850b2d6d5f6a9d32f78a1724b4f33bfed4ca2c513f464e84d8ed0c"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-en-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ffa1ced7709665f059ffffc43016e5d0f5038cb1f7c634af30a8d61b7f3f570b"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2a46fa38ad8f938e04737356d7110f8de4459a4c54e221e9fd38caea8092f376"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8b043bc422168a6b70806544ed461f09445a557574d2f21155054c55b47779af"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-zh-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c98c601eece4bb6c754cf2e8a7c0c87fa00cca3959e68f63e474bb028083f7a0"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "16f6906111d0540bd818ba1a1a892ff7c09db3a99a2a5070169dd2994d9e5a77"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c6dac79c6192bbaf91d052aea77640270745d32306529810405ff8e67758ba3f"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-es-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "246f467e411abddeba35d0756ab0773df928e64e70b81e4236e40b2b40a637a3"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5119c2ea9a0f77c97d35fbcfa0ad6f4407bf165c530d55d5f26f27ef260766f1"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7a0ee0c750b96c967ac9a80fc3234367797094d9526627f33a357c2723e5c6e8"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-fr-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "aa4158755cb22abe108eb775a24603842e1a20164903389e458ea87ac6c68ced"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8a7d13c5e9258a88e4e6be21a79ef37d5453053ab9ea76d507a6beb080843a38"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "447cc6fb3d2ada339b6a750cdd0b360f7e65d35175969b84eb99f02470298473"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-de-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "c3bfd66cb7bfc0fcd26e044540643c770a785a459100d7346ee5394c5cb1d337"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-AR", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d2f0f403a4c9832449785834d710ef89bc666e7c9e2c82f7c49b67d0c7356898"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-DIFFUSION", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "79ea3170667b29bd8a5b0fc54b6994d8bc665c6304de46b0ac41e841b5198432"} +{"id": "modality-0009691df35ef6f882ce3d99879bbe5a4d41ab14fe5b629a1a3d961f8b8139d4-ja-BOTH", "group_id": "diffusiondb-author-4a3f16da96c433731197f0dab70c6ec542331840e1c6174dba3cd705d5cdee10", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "aeaaf8ca25835bcf2f449740ba338bd0a9b36951e9e4452bd0b2a9c253579535"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "a3f6177e65afff64d8492a08ec400989f74777601340d08a23aa82c7fddaeb87"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3ebeda8075ba2776711f6649c2081ebf27c39bf5f9bb698ebb9dd0c426dc9ac8"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-en-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "abc6190a6c191105553dc78048f844ea6375c1d20d53b71603cf8817dd757b2e"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "713ca1303bb2fb075f3cc80881755527ca04f645ef72882593000193f166bde1"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b93ba0ae23fe45f34019eeded78f696b6c8dcc4ea86f5eb5a0e291aa5a28250e"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-zh-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "cca8d2d2e897894c850862e96dbe57580599bb191817dd6f2fc85599cec9b863"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "22c702a204735d981689f4c8a2581df2a8cfd24974a46d15148744f063e01daf"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "53d3deb80654d453c7ced4f2d663a674e54dde6a50b9c337893bcb221c02b0c7"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-es-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d5740d93faed0ca8be09466d69344eb8cc1d67f1f93b4d5d870ec3a3ee605b36"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c2cde5d2327d58fd0aee8faa8817dc194baef12bc50aebdcb895507ebf9cb21c"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0be6ab72e4fe3c66b8dddc4819e1eb8723ebe5c02cdfa13345e537bce3620b21"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-fr-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ac9f0e3debf7f69ad654627834be0daefc04d73987c0f18f3cc829b52cb43f44"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "7029e05250a3596514c932e223888d437a4fc114325fb4f04c5bdc73497fe426"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "92ba844cfd7f8919403d3c09e0ca811416c45e558191af6448110fb06922387a"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-de-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "ad37e5030a9eeed93299c8235bde67b3cfca648ba76b0a497db874e5cd07e7a8"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-AR", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "73e5dd2e692e2d9462187efe2110c63911ea0ead5b04586cce757d4c5cd1f94e"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-DIFFUSION", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0af4a87104b52fed5a08a0b5f9a1f84f8b6ee4f9230dc89585951903d4c8082a"} +{"id": "modality-000a059dcc007ab5dd97f45f6a5b566f3efb075d93d88fbb651b8ac9d59bc0da-ja-BOTH", "group_id": "diffusiondb-author-c88be0d3b910e49d9f869320f8ba3c4821442504c8141993c70451b3ef7c7a7f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f2bb9c87af56cb6b192d0b5f9cabcca327f991810612c64d508442a79e0d6c3c"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "20f3e66f1efed81164e7d73a9c2c4b6d4c206420edbadd9d2773849fdc09eaf4"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "1078bc6513295778a8a05865f8e2a5b596a6ed500a42de1f43fcb2379de8d87a"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-en-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "80d5e29c18b9b492b2ac1343238dc365fa90a847f59e0a4d529600283efc740d"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "cf5cbe7673e148bf2bc647e2e0912b22ae7db6d0171e75eaf7f1c7ee4884925b"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c089b3fef2eaa426965948acf7beadb0c5d4063de0eab6bb7efce5440aee68d9"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-zh-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f69bf7737a70cc2200f764de66b94656066b14b0455acd826b1a1c4d478ce7f1"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f641dfc21b534a6e25268d096e18507a40d1c7074b184d88a1fae90a0185d4bc"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e97972d66a4fd9651f0aaa088f4853340a9cc9f73b75ceb2eba83de871c77cc9"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-es-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1d6a23ead4b0ce4b6be17c2a12651cb358a461bb6681b2a7eabf2a80af5776f1"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "fcf33173086ff91d201a98b627330d58e48c1ffe2fda42f645582d9def645833"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a44c89e78e143698e0a389d7b36c87451ea8d0ec5e1dc4df5408da091bf6a5b6"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-fr-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "fa724e9a175da06cdb7d6e2c1aff5a4ba3d618d79154b0619bb41be47fd8d839"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8655943715eb421216082c1219ba27a56dc2f82ac6057cfb148ee4c61d279735"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4906fbe467d99820cc1c5f65de24c56e6c4fb80e2487633c378a951db01e0e51"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-de-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "121c8e963db6c575bf89be5fe8a48e15f82779269307956cd17ecde8a5f595dd"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-AR", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3cd6803a227d3b1017fb4045d09d64fdb9ef27b8160104e15b72a3ce175f0f8d"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-DIFFUSION", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "2e921777bf1681ec1f2de1f47899369f3307239f599179c1ba3ddbe5fd00afe4"} +{"id": "modality-000a261199e6fade10174af1dfd416169020df8be24c9f0061fd77d9a7703b05-ja-BOTH", "group_id": "diffusiondb-author-0c8c96e7e9fe91de66c1f3ca9c43eb973c410064f23c6352fb245da871a228fa", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "caf5278e863e37d569d58f1ace8c0e323ad93b6addf0056529fdfea95b0446ca"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ae329b752781b954ea098422d77c91509bec9c9c7b3df82d02ff0bdceac0046f"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "25265fe18dc6f29e8f704b3137dc370c19e2e330363671386fdf4b5dd1b0e7bc"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-en-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "0b886d589cb434a3e5673ea9be988a30d59035995650d7f8710362fa947265d9"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "e79d662e6093cb6342bc57879f9f49475417e5b872cd3a6a26bdba9cfbd216ea"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "007d728b76ca3f74d51b9ac566dfc5abe2091c119224bc1048db9791a8b52794"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-zh-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c9fcf52c3603fbaa60968d5a6b13b5045002a393159131d40fb1d34fc7225966"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "bbdc43f6fcc820ce7d410eb5c5e85e678c0c11139b32eb2a06b30d81b1ef0555"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "60ee3ea41078dfc076606fbaaf20c98161f9bc324683daab725131f366273357"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-es-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "8ca4c04a1be1c806f44ab4cc62c2a44c618e328cd09b5a8a4fb6c75e08ebcffb"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3a08f5efa04d598918fe44cacb913b7567386da41c3571244ca59349c0ef0812"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6c8055cb13e9f9ced5b77ba99b99928cf8f9e2f22eaacc31e804e51be524249f"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-fr-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f05541c283c34816bc398013e1bd668cf5259f920ee8fefed49513a638455a3d"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "7024b1091d05c12916d52f313bfca9e0fd1f70b1606f10857ccb42b6fec8d423"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8898f465488a9e4ad9ef213625f819a0ac4b6e30564966e9bd7a673a63cc9bb7"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-de-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8e3a4a3901868374daa7836b67d4bc0b726d702b4d4a271ee878ad2194fe09c8"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-AR", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f047566e8a9b90ef0d83839db811a4c02af5332c1b9926523ccc38d5daec9ab7"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-DIFFUSION", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8f53933ed15c76f5080f6f75739c859a9a0df1418479583404b7fea44e9a30f1"} +{"id": "modality-000a7376b497e975747bfc743497d300692be9455e2499e026b4d58c9492194e-ja-BOTH", "group_id": "diffusiondb-author-fc8e90aec60491db83556ed223f28b1dbcb46d65777f5eae8f91c2503491a308", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ab9b5482a1245f5760528ce67488ce853c42ce7c92837ff9616c1158b8884ee1"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6a49dbda20f48204c6e9f494d4c4fceaf95b23de8a4e0906ffbe2c8ca545c9b4"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d52aadc64ae420b3d7ca57d5db3eafba110c0256c9ae55dfc847cc9cde80fb6d"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-en-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "20c292e5742e833b26ce602b12d469fa71eba74cb4f9bced519a8872500afcd5"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "402daad24606a1ea9568012041a478e52ef9bf790695738ba509d95dcbc2c4aa"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f87729ec207cf50f2c3ecf00fd10ccaba7a730766423fcd6bc0df915b07038a5"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-zh-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "a57c6511be33a1994c5a25eb3bb77a37efe3f068be75d9299a22b92b9b3ff4cf"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9d8c4daf6930eb52d2d525dcd3deac0e18ac7f046e075e0c97468579d700f021"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "0ef2a8869ab0468f5f5688aab3299e6bdbfc2c2051b3f2ed4cc6d095276f6d69"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-es-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a2a47c7dbdb88145c24b0d6e5871e14082b313df4cdd69705ece595bb6d635a7"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ef87cb0d26d0b5bcfe6991249e05bb857285acd82cd377715a74b907d57cdacd"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "368f87a4888b18e3f89525dbd8796a74fc25c0c63721a885b6f310552a052ce9"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-fr-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f891842241805663287442cd5e027bd512473f602692c103b8e0105a47d19a3c"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e679ebc933a02c76b3d18bcd8684e98305b33ced178db97681b07c3afc7eaafb"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "5d65a2dd1e4130c22bd74a5c832d48561741bbbd1baa6f99d44b50e1fc888195"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-de-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "65372d270c56db11366eaf5fe8862820f256302ae9d274869f353430eda3dd34"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-AR", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "5257b5164ae527019ec0f7736bd88459ce61291db4f0c5e007cfa0be98fed9d2"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-DIFFUSION", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1a85c834c6b83c52db0299bd096c7ab157f13ac5ebdeefcaa60562be694d3f74"} +{"id": "modality-000b352be1db26dc5286a30014e59a67bf24aafe8ea0e71b726fc85b72133103-ja-BOTH", "group_id": "diffusiondb-author-6acf284e86524d3f79ea76b31b36a2dfe85a8cf9f6a2e38bd642ef6f13ab6b4f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f465bd77627fb50196c3f30661764d63f807344355867aa5b7e4562e5d683600"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "529f56124efa4ce6b38a63e3a0f10f76aeb1821acf9d5ff2490dbdb03649a49f"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "bbd439b6392875085b35ebc25d470c99444bdabea76a47bb6e4cec2584d6ce1a"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-en-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "4dee73f43f3dc76dd523c45562b066b7e14f11df3a21887d8de080c892ce2733"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "f9cf4ff6578589d8d1717d2636dd36db949331f2ea2a0acf942f0dfc928e6949"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1413bee03a52eb2c6d728474648da5c5f95c0ae2f7fae3bc8b5140358ca9b1af"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-zh-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "71cddc0852bb3407be0d153846aed74195b34d5a3f75b139b833f007cba94905"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f7b8784d51e9f0175326c09bbab6f58ad95b8128d55736979b5b2d12dab04385"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "595e69f48b557d0dd5a2120a1705a95696a6fe567b00d07febb8ee2aff3ac7ff"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-es-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "189706184b8935117cee39cec0833156f5a79195525e655d80bf9224c117a857"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6a320bb667d19f9d98f4d14c8ff90a37eb344efd439e4d32529d985c6e9e7652"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "25643c8d6c666ce9b1881a2d15f1074ca07bb42a200bbed70b70431c984598eb"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-fr-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6f85fbbbd1d5fcab451668c3d9f9e6154270142c285d48f228c2d6d5255d71c5"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "65e29a49d67c26677f63752294ed0b1b02ab621459242d803a15049accddc0ed"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9ef0467563f11a7fb786a750e93988d67969b120a6679fc853ae1df8283d013d"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-de-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1bea4758a9d1a2087552737e45d132e31d7923f179684bb676b0d036c2663293"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-AR", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "cedb52ec64a5ba7c359bea48375c582d37fea3045f4c5ff88c9111526eaa7922"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-DIFFUSION", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0bea3b79026dfb474a9b87ad851acc36123aa85edbdf1c71786f27d4a8ec371d"} +{"id": "modality-000b4ede20f84e53679b4ef9762b7fea7f23435f104dad2ef7476e0ec4bc4572-ja-BOTH", "group_id": "diffusiondb-author-f2191195d4f8cb03514e212ea80f11582abefa65b656dd9450f374cd01e2f49d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "585e3ca2be3cf83869eabfd89ef74b55a99683b30504e5c444621f29020a9463"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ffd47f37ad691a151787f5a760f801190ff60b1e1ecacef0152127ad861a84a0"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "67453a91f9d69ee4a014cafa539497c38beb5885abb59426a75ec403702bc59b"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-en-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "5cb2643325caadb559e86745335ea0dcd43565a966302381875b48edf612498a"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "e9261ab39cede4afc43b5188180b858014ed2ae991bfa2098982917192fa595f"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "97633cedbc7b6990a0dcfc868377801cd95950636472a4f313f64c885d347da7"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-zh-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0d77229864c9ebceaf3a0deb224aa93a519787e0eda719a5d4f935aec35c1661"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "01416c994fda8da65bed203232cf3891496d20b8bcf29c602cb3c33f9786927c"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ecb5e9feb1e381c9d2141d015ae1c329972e55ed6a5b3c9976e6af4381545a49"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-es-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "528cae077cf55a228ec323a17827b7a774536ef571160f57c4b2051245bacd5d"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6ffbc8c6188bc25f49b1eaad9604401b316118f270606dcbee46df0cd9d5bdb2"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "bf18bfd22707a5d5e8a26c674f7d7c4e0a00b8c1b4c339d90e55abcb330f37ae"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-fr-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "d242cc95ce86d5c008ca6adb40e4a24ee55b3ba1a48c01becd208e2ea08414cc"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "36184cd03d35e05dd42c7c9a57f528d06e2139eab62582dd46e49c33c1bbfdad"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "de71f8aab1b63df80c320c6fcaf1a783ab2ea32d99a4a6f4ad7d65dbcb473fe3"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-de-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f47fb5bed89bc965d1213ee4644394608513261d09736d097f4498e070caacc3"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-AR", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "099948de8bfb71c461bfeb210cdc2eecbee694669bf1b3ea84eadca3b50d8459"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-DIFFUSION", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6bdade9e63ce04a9e0f3cf86001138bc9b7d8e5473a6e6875ec5e399bf7663b9"} +{"id": "modality-000c4b75867badd51f030132f158a072d6fa913a1238ce0ba55550772143e730-ja-BOTH", "group_id": "diffusiondb-author-5baf7cfc60e6fe5ad202fa357ec3af08bb74c8dc6dac77f3f774a9b35695dc2c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "122b349ee7181af80d36c75629213e84464851d28803e94c79bff3bfe0095cd1"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "48f3d42e804e1688ea3b0894e7a9126c15c432adf15705c847d1d9f6d3f2f2f3"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "4847b597498a4388314d1eff391b7b646b76f590481b4ee11ef58967f57b0b32"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-en-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "fe738db5631aa4f066505179ef3463e50d47596ab40e51fa87af15ba1439b8c1"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "dcadb7c450d4f13466e32ea091904ec96687e7998fec6cd1bb5b524deb7e98a8"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "7e7186bedd1369c81922fb4655ea95bc17faeef2cf1d842b7e7fa55ae2ce5657"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-zh-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "facd7d6d4ca536c02b93a661672a15a3ab0480adfa9f6191e20cf2037216fa1d"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6cd743311c712c104fd2b4b393231788da2a46ccb318f5403dfdf11255d37dff"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "0a1161236be9be6d80aa9bc8f4ca500c8224587355f18a642c8b3295fd0e4b13"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-es-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6975e57eba5f705de82c4427001f69d2d18e582043c9e51f5a364acabb894c95"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5276919239bc4efa267d339cab7b50a301fdf1fc922de0f1f56992825939d48e"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "95b520273d9ca17ee852393c3efb8c516eaf3a4a7a063f9c001b193a02709150"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-fr-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "a54aa138b04a9ba5a5da63ef2f4fffbc35ba202eb1c5d46bff8d63d602a0cc9f"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "749ae7fc4e44d069e6f86360156e8c86a0d5c38fa41dcbbba992685b0345f297"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e9d077ee85d4dbe31e17db7d538094f04b4140e149574770d699088521e88083"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-de-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e4c4b0e4c41366bb4c18579b9a6be224784e9f48049c7280fd1cdcd1f540a425"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-AR", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3e2053819e7f4119cfecf3093b409e5e7933092d0f428fb71612e44ac5893505"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-DIFFUSION", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d39788f2b281d1fe25a96edfb7b4a7fd3509e8ca296ea5a1706a99e92af325c0"} +{"id": "modality-000cf1ab6dc3571e8b74b0c7af95ce2f4849d51bf6806e75e792008f9906f5c9-ja-BOTH", "group_id": "diffusiondb-author-7798f6488acd86f9c36e8ee02e8fc5d209adcaac4d997bd7de4ce73b553b7f9a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d6a3ce5645710f581c84fdf33fc5f90f4446927a76e7eee175a32e9eaf7bf07b"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "cc653f50f5cfe384e21ed9915256451dcbfc181f60ec9959b2dab4177685065e"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "842c2f4837c8732cb8d9e38cb94aa0559e520f777a50a411ce1830afd05fa0c0"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-en-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b899379a5c620fc6b8e81f04b3f15a671d662a3482e442ac728d6ab7c4c35126"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "7d66b76e83a4100a9ece16e489ee7483871950486d48e5c0a018448b5455e8c7"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "38c4ee894e4d74fb965326bc82de98cb775b280ec2c4a4adfc672878fdb61b19"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-zh-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "640181da029947f403fe3e3df1e50c7448e2de1ff265e80c3a2a69a44e3a3f84"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "58f69b9f15e305a1fb6d6cc4d08afd2f2ca3be307afd4f405a13881dd2697b83"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "dba695da9065c9fedefa28785713ff18415549de2fb057d88ab51c42e94cb817"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-es-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "caec809dcace5286c84756258d52689517b59d064eca714d2a0ee4ea30843624"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "0a19fdcf6afcca7fe9ea86f5a4b3b109f885dcd34262f361e26debe434afc991"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "07009b912db2cbe1e278457521483d1cb191ef840cef0bba6ba83d241544f8fd"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-fr-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "1e650cd47ea5ac8fec05eca3b80a1b3921bd29c72403c255764b257f0f6cca6c"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4e1143c2727c9cd5c9155959fb50e3f220ec166e289e642da596e59445abc1d6"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "bb24121a80d11f0568e00a7f33c76aa7674b2d19de050e31e187fb0ca8051fec"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-de-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "02e0080c0955f7f205975031f6bdbacf11154bb92f9308db0ca5f4f9c8a2d8a5"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-AR", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "2204b649f8c7d45d8787d92eaee057fd15ab7e5a78fed8aea95a511670daf26c"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-DIFFUSION", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "5a39457331e9aad81b287f36b52f83db8935e27a3f0764acc14e57f1659443cb"} +{"id": "modality-000d70607e1baada0493f9f6c0c1fbe7d642d087ab9e14fcbb2b9a5ae72d89b2-ja-BOTH", "group_id": "diffusiondb-author-f2fdacee496b987fb5015f5e64393c5caaf6b1f50f49ec5e5063aa79e56e4848", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4c10f01c33511f1e86e036a6d05e4b497175aee7384af96a55ec7a9a0951e1e0"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "7abe889ad9b4e7b5c87e3a297f2b106b5c1ed8e626190b606ae58b82d5887ba1"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d4f3a2f5bc68bbb0e860ea859c9eb7aeb857345117f3f92d9a348b56ece3f28d"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-en-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "aa696983fa7ca1462c68b81f8f38677ac8a93aeef7663475168442a238f5b36c"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d0c9841b82f5a78fb136a916ea2b9a1511fd44d46fe8d1356fb7bb544c3785d8"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "85999dd9b42c02ba88c2e05e930b64a4ea05d939a1551aeb32133c4aa4d3746b"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-zh-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "506371aff8f83d3e9e305afce41f1722ea217ff6965dae925b387db4b537d517"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1c52d5ae7f0c106e1d40fe9ef7dfba8f5dbd6bdfc3e0ea517b8431e224620fde"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "40a26baa1ebdf0519febcf683c97f61f56708a5119dd7b3ef7918c917d5c8480"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-es-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f1da3ede0e821594b781f0decfa2d9b31b183df7c1aac22a63fe889bbd7ff973"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e46d0214ea2e032b307faadfcfc60f38e814b9291f40cdb37a4595531226e7ca"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "9fc4ba0849ad753174158db7440662b1ac129b2c2dd1be94ef070229d1f16b94"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-fr-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b4e57d04291fe7bfb2820d3f6ac1859d40ee6d0d642a18011e84a53c1690f02c"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "348801e3e18e57c711d289a83e2edb2e5d212245c1b9f526097a5d264502d9b8"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "deb3800ea051d3cbfb661b789d56efeee9ee38b20122f53f92d4d8a7e798d317"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-de-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8d45d656dcb5aed04c4fcc8864bea9661fbefeb6f69578091141a46753bbf7ce"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-AR", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "7ed7b0b6fd5ee47769cb33b344e17c29104f7331bccf7823ff11af740937495d"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-DIFFUSION", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "380681ae702b10817b75b6db9205043d3e39fdb4485db391ee167066bf4f453c"} +{"id": "modality-000ded134794c4672016b5a8a76c4665c8cc501ba611c1d28fecbb3cbebc91bd-ja-BOTH", "group_id": "diffusiondb-author-3265dd43a5d8c9a79f9bae56e127c32afb0a1accaae1a454e361df53cb0462fd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8d80a91138ad9022361af37abfe49ef3079acbfaf041ccc65f6a8a288a76e8d5"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "c2c028eb173081e87fe5cbfbedeac60e4fe521d3ad31bb235b2f258d5491eef0"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d1d848abcbc36ad5d6b084bfbe3b0254e63971257b65cc138341cd04356b86fe"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-en-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "87f76f0ad4b8b7b86e14827bba17a361984184df41877d942916dc703b5f2793"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "66c6af6464620245122d2774a2d71714d1df0f3999fdae6eac3baa087cbc0410"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "45b754c6aeb46104337e7fe05f3976e361df9fe57c4121a1064f886f1bc82445"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-zh-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2eca15310e547b03824cf713c64186e46676bccfaf740ed3c59135ad3c0ec7b2"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "2f9d7b7035b0637c7800004c0170793e5c8b181fb4f430cb9d89a96358ff2f51"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6238467a7ad88b73f9e712fa0c523fb96df710de4d90f0d0dc2cbbc89b2de14d"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-es-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9d7a07e62d6e39f80d27a0a643f18c8fb2560a896657b1aa73a3efd1c0308110"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b08b2e20ad7280c8c40a12eae67e5d193e7261a38295eb8e6c70498ec6c7a4ba"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4944d9074e9030dd33927af3d3fad40255d2f9264fe84ed348ebcbe6dcca80a6"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-fr-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c5da64caba60b1e808142c2e07b5732eeeb4a474a7f8eb72dc489bbdea28b3a2"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "133b5009846d860e46340d89e7ff923b847c08e7e44325fb433a410c1e271f35"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1e6a7831986fc9fb01477f2524afd51e34c87d9755849fe5bb7ad7398ccf78e1"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-de-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "a254b5b2d728811bf50ce071edea1a2c0e9413ad2136f5839281280f586f0f17"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-AR", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6060ec6856b6da52324184f77d8ff9e78a6aff927665d8deef85b771cf3160a6"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-DIFFUSION", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9cc856922ee696eac192395c25a84c53007d504c3322edd8f530f8ced458d35e"} +{"id": "modality-000e2daa66660c1c82e8b80222931bef9186ecbc3734fd0d7a94850f41bd0b02-ja-BOTH", "group_id": "diffusiondb-author-569f78b7e6ddbc7e1eb4bd3b50a93851551ce0ddfb2febdfc5e3096218d813ff", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "03633578c418db0379a31dec6b1326dd1f4b8abec9bd6d70514333d2cb0ba112"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "654a76c0e305b03b66ff8784654710f69410dbabc174742f0938e56c851c7636"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "30c484cca646fa876ca18050db55845d6420450764060d5d42c10168f022c855"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-en-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6bec45b8aaebef3dd1046a27858519c28d7a1f569abb132a60f4bd120e3d70bc"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "7025ae59da5b3e32d3460aad4d7aeb4e52a5aaf3710b998dff6ba7049f27785c"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "fddc0d8a0e4d8b9f89f9686da2d8e5a92954f8a77b4ab7afbc70656a6be21d19"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-zh-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "31937d7391aee80abf15fe925907970224d17010de139c0a6532e4a531530225"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "be2a617e8ace8002caf8fa4d2ddac11251375ccc251817b2fe1e4ff30360b0b0"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4737cff6bdc8ccd610c7825af098479f4790b6dc20b61a5fa533d8c208365dfa"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-es-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1fe0d6883d10b1fbea50958c73cf821aa749194851c65940fe98f917f9bba4d4"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e77fea141ebbe6872f866fdcd532e8f153b9b9867d46550bc0add741965b045a"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4f5d22c3b4b219557568900a1ed5c12ed3beb063856f682c413a5a12416b03b2"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-fr-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e2ff4868823808442e14630978aa7552a4621cf7937cd89bd26b426495f7c38a"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "41fbce8ad006ce73746173628d6c2c9b41443614c1fae2d14aa3b1bf9f889ac3"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "22da016ba6be0a8dba2148cf6b6e731320a0bc571005376891b1ec5b75095c77"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-de-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "5fd8fc653173433417daa803fcdc4d359262a83d0ae057c8b7c887782c7b5ff2"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-AR", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9eba98d2a8ccf40be85266520e21737f63a211b7da927b4ca64127259004e779"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-DIFFUSION", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4e6314029c0ae91bdfdbc00201367a69567c9674c235f99bc93d3fb01ac24318"} +{"id": "modality-000e444b6c3a69eef87c66b67718ba305a07bd6df33306bf83764dc8f30f7474-ja-BOTH", "group_id": "diffusiondb-author-b5d0885b21df200a1fb1665ff0524fff2353f8db50795a14ca0c30a2175113bc", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "652d838d7e452ea771e2ec36fe8a891c23d370e5566de0ad8e4305faa6149834"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6f7307e9351a2322e2f8404f016a1ff34f1a3383f744c2e5d80e3b497aeefb09"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "7fc49b2adc5f4b7d8339e3eab8b730b63f310bdd8d9c48d9143b692f718f80c6"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-en-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "5a0b0dda3dd45b9ad2fbf361d870cb8e95a6efbfaf9cd7306c2360f4fdb134c2"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "578eb9070a59121bb21ba667078ebc13c284ae1043f38cebc4bd1d915d3c334f"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "77db11501450459cbb32494ab16c70f1d8484fceaa2365f416ec451e9c27aed1"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-zh-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c0e4afeeb21e163d43d82d05809f2009956c564871e5e22a2f48e26ff2e133ad"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9daca383d4a0f11316750495df0256e64969ccfb5f44371d80935e2a89cfb90d"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "afeb6900d5207f0909573d9eff1ac5cc0e22b262fd9cbed3dcb6743158fe1f63"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-es-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e18e5e2e4b8c59c0a6884c7c9b18a05c802384b2e12d82e619d61392e58e664b"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "901e55872fc37f276fc9fe80b832268f7378bc432682e0d5505ca8bdb36e9293"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "bb84022887304e6084ef74497b56c80bd6ee3f4168f35426f6b092d6eeac8d69"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-fr-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "bd2c2c41b37fa482581b97a28c25e0f28d602953388a425150c5808b0a4c7073"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "2e7283e823c1d285094c733ec1528764a9e871220afc92d76b800d1ed6360907"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e604604196512248acfa51705e7f0f67eadefac7b5c1e2783d07fb617f92c9d9"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-de-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "645368f8bcd952927ac1df4922316c4070d4f4daa0a71bfd76923d8d23a188cf"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "c1ef33e0b8683c5a8e7980316d3bc7b9fd01abb99dd04752f4f71e161ba4410d"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "43058eedb57e99ba2c70de74808275cc00579b4764b38857f98e4e5794d54edb"} +{"id": "modality-000ed006c10135d2f1dd1b6229bbab3de28211dfac9d7552a37a2de8db139ecc-ja-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8f91bef7c3905bbdb0638e9fd90c2336eb8e3f7f88ed3a8d822434d2c7645579"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "f32e53852fbcdc6e4d575024f3edea059950a415ac8d2d6f3f13aef6470c1c37"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "1f770805cea1a619abe600b1e1a3c5e733dc4295c9709a3522679e960248d59d"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-en-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ac5da0e21afc312bfcdf0df66d0a6776b3aeeb94134ac1f3097bd8d0b161c34e"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "60e3545dc07373547b30babc88e14e52c0df20af8a1a6b78ef0f085a0bd22d57"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "5f3f0ec0c081c5670a93fc8a64ba99a2189b672ac49e42fe9825a7e544e9cea0"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-zh-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c0cfe437d9977cc37fa633c2af7a1dc9461a5c94bae29cd1999215fa32378cde"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "be8e60660b45cef8fe88968f131127b2c4b374b99982e99b2e4cf03ae5b60dbe"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "b3c9b03053ffcc305c6a83055bcb8f46ca4bfbe61efc66f875c9b1ef60a96fdd"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-es-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ab0d9a55f2656ff66289fc952b160f8d3a6593c4446f86d7cf447a3e45970dab"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "6402dbe0770a617567ac1327f22a50a970d8d69bd80566e81bc36cc85ab8b1b0"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3739a1f323914ee8a6f40d75f72f891e2338ce2447ab2ef00c4eeed00ee4a67e"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-fr-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "00982c65880334c5a8838403089191781d3f0ba8f6c0e4a2ed346d2ba6fabd5f"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6ecbb367101d98e062e4abbe8f968342b4b02a8a0d7b60c5ca0f836d995046d1"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9af168b3ebfa53b9d5cdff120ea7950bb9b417bf9abea9e071df2d3c05d4c2ab"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-de-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e34a9bda58e020c2d0dfaa40f329945f56f39d7ff94d4cdfe93fa48ef05c6128"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-AR", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "69b4204e37c42526d9b371d90d7b72e5a43e640c01d76b268ca57240e0529d4c"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-DIFFUSION", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9e3ced1b752969fb166026ba018fe14dbc296bb71129d94f165c990b438b9a5e"} +{"id": "modality-000f2ea9dc48eaa6d035e3fd671f1f6c4f43aa4b4c47ec1cf78dd4bbc5a7aa81-ja-BOTH", "group_id": "diffusiondb-author-dbed259054c83808923122da3408581737d725708962cf24ceb89d156c7bfe0f", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3f5fd7967d389ee0d87fd1adf5217fcd14835f59293d9fad6abc925b89e73f94"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "bb72f8f5a9f047350388d02444e8aa52f9a569c1a8c62e8cd8c3e22e655afe63"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "be00bbdd0d8c1c09e38abed1e5ce8ebb78af3f8a460b391caacc667281a7f60d"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-en-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "1c7da1347d19844e97994e52c59f8c43732511288d859c28fbd9e4fb7c085c83"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "7bb8d52a9ca5fed76d4f123c55ffa3a5d1f030fa7c116b431bd3da1e474e8831"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "828d53cd1ec2cea74e7a573fa238a3592ac1f847231dae5cc5b116640f0f1270"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-zh-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "52352f346a84a4748720d9c232c0282e596999e34d03289f47ba3fea5265ff6a"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1d38f56cb6dbe78bc5ed40d85bc6307a97e2fb85800d14c90910dbf53491795a"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "16da5b9a2b9961926d9c77e20f2e2411393013cd06bd300f6e02c1acf466a36f"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-es-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "259b860b82398643b1898b3232399384ed21c4cfa2a2b7078b8ef03388578520"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "efce65a9422d87004fc42194dfdf32c30d404c09b3cba3dba95cf915ba7d2082"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "40e6687932d32cffcb688d70814b891ad37b0f2596620543e4a82eabcfc9d7e3"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-fr-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e6079502f0bacce6b459e33caf610207781c575a082df2f70ad914884be709b3"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "7a64b004e867960ba2fb10d0a9d049487cfe12bfd92cb7c0e610b507fc1b4ce7"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9ad6f3774144f00678911934c5195bd9200f32239e96b5bdf86f907243c260fe"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-de-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "920516c4ee4f2f160126dc34e1768b40a71e62060a3378569fcb6c6cf2e24767"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-AR", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b231dda9b3264856153e28d82cb52a5f794fd5e24bbe539a16db7d295fd16d53"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-DIFFUSION", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4b9095d72e326392b967449d62f590f39e8faa62810c208efb6ec0913bc32b45"} +{"id": "modality-000f7483bdd6aa84d1d445b1bf116c556f3ec8a45aeb092f19540712b7ca35b5-ja-BOTH", "group_id": "diffusiondb-author-60108b5f1e611a36573896d06f945cea6507336af6fd0030fe775e6374c40c6a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "23f17d7d921308716dd385bb0f894ec95b57a9bfd691ada63d18b48c3d98ebc2"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "be6e914d284cf0ef2622f59a18de3124b9a565e766250f02d3a2e791b6d84f27"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "682bf2fae79ea71957de4d2c8a127e4756238f0a74f465c279569ced9a45fce6"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-en-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "8fae335cca867dbb471b456226515dc6c60a1bae345cbe8b35644ed648687cba"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "edc204ecf5fbd6aec5f32342792ce83fea88bd7ad6f4c582b6177c5b304c56d2"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "a60b9bfbe21639fef8af4208e0b48bbaf6568fd80f678e47297eff5282a39561"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-zh-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2323e915f83ab85db4ed9aec7d35aa777011903b1f1d3fef5ba3ea1afb1df2e0"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "20bad20994cbf12885a3216937219a2c232db71f31df0064e5b64652ba874934"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "5618e8268d0508d2c526a7f24d14d3b686860e87dd0e61f523a575b715809463"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-es-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e3a901e2f093805dba7bdc87ee6e492e9a7926d8241b17496226f65bb3056f54"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "14ed6ed23f652a0ab16bcc3d98b36b2528ea7efc99cc50831c54b89f51eb1201"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "621803defa4ffc0db37c0d2daccae588ce5b90f4317d28aab6c9c9841b3940cb"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-fr-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "11072bb451bc795f59c614a00368b4331768481f82492e413381b3f6ce2e289d"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "9cf28ddca0e0659a09d7415d45fbc6d0011abd0aa31f0bb567aa1a0f59bc40ca"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "db1f5f94a28334c17fb00771785667a43dab291b0a64e14a046752d41618e198"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-de-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "bc5b887918c20e464011c5741b8ac077164a264dc90a525dc801d4ac28fe3870"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-AR", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ecf38ac2769d2d5bdc7e8283ffe032226092db626f4a504b5b9cf2bd8d632f4e"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-DIFFUSION", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "339fb0461bda42aa9dfa4f9a18a759bbdc473b45f4ff6ae0311796e5dde7ff05"} +{"id": "modality-000f779bff858f83a8d5be3363596289a64b16011f5fb3fb50a3c7ce59834360-ja-BOTH", "group_id": "diffusiondb-author-a1c5d7c60d0b66d1c8990d9abee1829450ab3b0469a061c323b1b00e55272404", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "009c3f018ede03ccc3b175d98a459a9e2483aa9db364eb070a860e99c2873acb"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "82979f92c660ac11815641b32a5a323b61febf27e738fc5fb2023398b1bfe32f"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "8eacd984f2d5f795f2a7674162443049ab3129c1a07d95b7138abe46fea51e28"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-en-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "64f2b5ca771d7a022ee59dbac8efdcce20b3c842b02ed09b7ef4fd79415d3308"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "5ecc5c9907e4958ee32ce6aef54ddff85352af04af49cfc27601ee800fda2120"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "9c154ea427d09d9ef9bef17eb51928896b0a625c451c20db48f46e87ec2e921f"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-zh-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "62038ae689e48c720ce11a31eb26a2ac3a45841778b90d7581a3a69b8fe81bf1"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ab824fbfa8d18943f51000008ec21fbf173b7c2b85a30f7b27618db28c20c89e"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "31570d1552be101f4a4d833b086fcabc6ad0de43e19ba2c43b51781050f09396"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-es-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9753e037e8560423195e2aef966d4cd7fddf4494bb3d07b2004084772787cbd9"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "5702edebf2d618426f89f5d28c8d041c063bfc0be5c419ee8624103dda342d20"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c6908f4812af695791e08108c757cc70c1497b2f4e4c74c61df15ce25efadd35"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-fr-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "41b3d25222b4a80bd2a88c38daf34a5f3d4bf03adf804d0b4ec7d35001163682"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f73c43f0f01ee2fc94330c6677558cc4496b7a4d2e82e180443c7f1ba282ce4f"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "51f62860172bfe1425b9ab1a2b2758c91c393b384a06eae5ee080540647001ab"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-de-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "991a24fe4f8140c0015fc4688d65e4cd927766abbd575bd94985fac0e3ad4e19"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-AR", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d1b22ef6b94f6b0c1719024eba1e781e90faa0e94199463ee2fe36a6deb1acdb"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-DIFFUSION", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "14f17fbd4ef6462bd06ab19186fa718ba6640d92b5a82229cb9a6a96dbb9ca45"} +{"id": "modality-000fc0c047531d86e64842eb83889bce0335827de67f234d0e98bdc6f2af0554-ja-BOTH", "group_id": "diffusiondb-author-d8749704597157450c7a9376c80645263266b2330736a44e2a67870b4846c359", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3414e22c219d10b0fb3fc4dfc6294c3b5d1ef015b3f1f34ca47a2297425e235a"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "45f3d5531e6ec42e86da04f1ba48d1818ad5a712be70eb66b22a6a9d4a05dd7b"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e1afd43755638729cc5ab7a7e4843f551204ae151077dbc643816b21edbf9a7c"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ae9da483710abac72de7461c27d783747f079e3549a51c653c895b97b3bb4f2d"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1d0c62b63ab22da0fa4ab039fea875ef551ef9de1388b4428177839973fdb7fb"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "acc96a6709717bade1fa9f5ded14583083335fe69f37e7298b5e7b0287d9bf8a"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-zh-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0cf787988ed8632225625600e0c0cc8737ef67be302ea639069a9d8cafeaad7d"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4f94b768b5e91be2eb4beef84a4c9a2db7ea7b29bfcb08f8123f457b1bc825cf"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "aaee2b17edbebff5824bfa9c893f8dadc7101dae5188c25156836cd32f530646"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-es-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f846716b64b8f26feb2a30b6aa33636ddf356449d2d2f2bb8c3a33bca9dcbe76"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "9055f268b497a535718809f6dfff8f4a954aac5519ae2ce7111da0a4cf7d74c2"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "77158cbfd5e97f8387fede2b0f99e713cbc5d581afe1c35f320b90fecc5b21cf"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-fr-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7f47b43da058b984455f5a77376b6d9733281d11db9be6646011fb6243a829fc"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8c2de3fa66d79227e9a5bd28b6574402f87f6e76c9bcdb2eef6359852a09bc9e"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "89908de04f4bed96480fad663bc8beb7b3b019d6f7fa59e00227b8f01c1d9791"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-de-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "76291d52caa246b620947ceb4042fb424d0e635cad602e66fdd8abed23c2a491"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-AR", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "fc255985b413bffc7b0bda723d29cb1e09e644e089684c2a47538c22bb16ff44"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-DIFFUSION", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1192e5c16444e9149a955151fcadd7b997382407ca69940d79b4e609591dc2a1"} +{"id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-ja-BOTH", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "675a31aec9c98a0af3204e6423aea6da5a87094c27b9676614589bda17bc9fda"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "429070e643d6de748d67b8ef809c8a2dcd7cfc8f542170a7d3d0136a716ceac1"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "82e391a191b46347bd0930126eb60c2298d47079d87e0b9f0d9d7f8440ec5e18"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-en-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "7a32cc908dc25f9e64296212be02aae6247a1d3f12e25d2f5164d2bfba680220"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "bd60eeeee6c1e45ab9095b1ac2eef0cc0855355891528aceafb667fdf959b2d5"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1ac6da22d3ac53ea7241374031accb6bcec19391c91ccb60df814512b2913abc"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-zh-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "30e4c895daa608c465d3348985ec45a05a9fd00c6860beaf49aa3b8a909ed0e2"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f555588460820d154bd7c52782dea73128687f3f7f4749d2538a105d5007bb50"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "83b365dfd5a7bb71b60c5b69a26455c66e42b831774fbbce61a809da6f504990"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-es-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9f3411be7e6ff132b82620e2a1c1c8d9529030414e1fb0e948be252f8ee445cf"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4c3c990e59fc86b68b00ea8d391ad5b494fe4a61a4ade1c720d95cd89f6cd2fe"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "52386795a9bee41d7eee5261eb6fd41eeea84426413277bdd6d1579368ab6eb4"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-fr-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "85e45349134f46543cd834a68027961b89bb15022e4320bdd4722057afa2fedf"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "a76f8c1017a588ddc2dbdc02137e041d02051f98dbe898d5966742ab101e07e5"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4e10f0ef7d9b4d9752aa30b0bf174fd15bc277c4ed72274fd76473c361f20a7c"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-de-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "b00ee4d0a7b160d84ca445b534e33cd9eb4295b7585b54010ea297ff1b74355c"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-AR", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8d2233ec493b54121a8915094728549fab0dda18342c7465be479ec41806b6b3"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-DIFFUSION", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "716d98f18b2b0e28a87de6ef02b6290bca05078a26642ceb479cca07611bf00c"} +{"id": "modality-000fead0f933cd78bf387c13106dc514302f86c4e15070d54d4348f2272148e1-ja-BOTH", "group_id": "diffusiondb-author-607017b23135b1406739c53b3d60dde86136a308810b4e7782aee6fc345979db", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0b64816d577777396ff60e2282554ce7ea505d273b382f2634049b9d61ce4b84"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "e7e063231e61f347c944c87510775478cbc85bc7bc80eb728609b95d7778fc69"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "5c52a9b624e0757102d4bcaafee0a9cc92cb4ed8c546288210b83fea67505a49"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-en-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ae2e456a5b004fb2f3e4c5359972a85dc9da62f995db645d6a991157d84bb329"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "9e133d3a4ba36170142ddd533ba0dfe66b009c92d63a8ed1c6ced9fbcc012f30"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d6731a36ae00bc968863a404e97b9728bf57d3ec66e1e39f0c121c8697ace62c"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-zh-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "522b9dd9fdced83b64fed3a5fb58344395f1d37df664998943580055b11cffc4"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9fe19c208250c1b01650df88115cc0a7701f33c8ea99235b141169b0eb867d73"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "7c4470097888098810837912f8e72fe4a0deba412ac918111beff30ba907edef"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-es-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f51e552210817d05b7fc5c021a121874eaee9afcaebcb6d8a3d3c6a0a615e6e7"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "86690f0cca85b04e4ffa1b994001528ad725ac886c09eac8813cd8d06307aa4f"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b0f9b16ec1133579eaad5493d6a4bcdb7a6f641605dfa4bd97b04d340bf9b3c9"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-fr-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "c4980797b21e7f272b6eca9788b82e119befe5f1ecdab3509624d4777e1a8b80"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "4b0b54c1d379f8a384e96b170ba118c719bb67ff322edba3c0d56074579d0487"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8493380a1a9eaaf58eb1c478bb1d0b9b481792271b10cf6f1ba03f10d60d1852"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-de-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "74b9837141179f06e11b834835c25af2596cc317b28cfcabda0d99f43acdc12c"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-AR", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "379a06ea49d77eed14b54005857074e42b37efe17712bda2ec0fef88c3c190ee"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-DIFFUSION", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3c327e4354c544e93f9351551a48f7bf61cd2e6548fe8a621db048f8b539910a"} +{"id": "modality-000ffb106a9a7ac4acfc615372364101dedec59c7575b2b54c97a740c73d43e0-ja-BOTH", "group_id": "diffusiondb-author-39e76034ffacb254f4fa3d3619db90d0367854221e9eab29c9414e0703738a68", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "8e16d514b534bff256cb47d27cf22fcec4a3487e51345dacee5bdb40986b47de"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6087da37c1e9feba97f2d6536cd30635c3b1f18bbcac89bfd8559a00d430412e"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "39b2bb0f9cb9eeb810483d3cabd20032a8c6927a12f7216c46c34e5ddc6af2df"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-en-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "35efe5914012f2070e9520bdbe763d277625338339948e6e031523f7aa85c513"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0af6a082b271cc7852fc352d5a754bb7830cc732a2c66cf7a078d8431aff2703"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "22e263119c8f18518b6337cdde70bf58c5297c091125d3c394d310d7e9e6436d"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-zh-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1ebc921c39b023d41a0aee0ab69480becdcbe413e3a2a775c7685bd2291424fa"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a8854de99eb428b3909ecf809801fc0e129b7d8d9031782f14ebad965623890d"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d09754842a3b17a36b2dac8cbfdd39762551bdc4c66c072bd537f7a77cbd40e9"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-es-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6fdf7905c4ae4d6b5b3cbe9493114df9b7e2fc4e8fec0c7beaa9c6d2d4c788ba"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "14db7ddd6808d6f44678113ae931d8d901836e52791481589feff0a7259f890c"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ce6946d8cf1f1d86e1e40126cf519af173ea042a5e500c00c755b310648b2a75"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-fr-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "bc8316ba7e8cb93e47b0ef280c15afcb717f1f049b7b81c2fab4ef7fb0d201d6"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1c04692df138efa2fbe80aa5a921963053d7750f7615ddc8d447c895043c2a32"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "df54942ca06fe13324fc09b7d9760112b510146a7bad777062ada6fc372c012f"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-de-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0b34a193edbcf2bdcb93fee6c24d216b27e5a1a7bc6659db4aab4fd21d05136a"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-AR", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "30da60d687fba010b951a24a2e43f75b99d541793522927aea076b06d4355ca9"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-DIFFUSION", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "4b2310fb984e7e9d338f137b6cf81597c8676a6ac9c0c2209e7b65266e0b5513"} +{"id": "modality-00108e5c397d35a757eb2330827a105f1f65ab0443ece49014e0bbdf5b7a5910-ja-BOTH", "group_id": "diffusiondb-author-9d8d1537f217db925f131a26e7d32f88304a3b32958c24977fd3f73cf875ce6c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b7f6b50a2cd13836bab2413b28cb03a39749a83d86581c01dd4caa5c0a08c96d"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "57bd6530c35600ee31e4b9c97f414c8b433280a0f7c8ae87d4c336eba6b7bb15"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "f407daa17f6618c8d8c0a33a3ed28286989a0e75bba6b2d9a6a7995ce35e0d99"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-en-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "47fc981644165ddaae5520df4bf14cb21f5bbf0caf8e3a8e1e54e98fe16474e5"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0c8917e92c14dd158f58f8f0bc8c6130de07f77a2c102000a536f3f574cf8fc8"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "c98db876a55b3ff045be87187f116910896bfdd698ae2c499dafe2949ce9ebb3"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-zh-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d41935348032226bc29eeeb69bf8f78f2a396c63f8233de62b0e40c10c004c0b"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "de7e51b233657dcc3c1396950da314b4e2bf2b782ed359b8e28e5f2385e05e27"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6be72622706d6e42d2a29c5c4ce9dffb0901e648e31fd21b0d893f3988374bea"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-es-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "fd2dcf7fe1fa089cbf6829f6eea8634c333a08452f8ebaa7b16e3aa66b6da486"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "744813a0fe51f66dbb6b9912a6b7315577881be0fb285f8cfa777cc068598dd8"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4217fce9a305188c6b0aeb1d50dc9f55813a12df541c0339f63fb5216d039380"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-fr-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "7cb0d8577009a8fc56c650a286cf5336b8fc92f77e8a23a28e85377654b7f385"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e8b90c5bfbd9daa0f64966d83765e0ea6694c41f6c9c63a8625712b317c13203"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "55854b55be769cc61f74068e6ad73a39b1cb738f9574040c165312d25a149de3"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-de-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "6ad7018d83748c541e0faf93031986c665a1aed7a24740812d9be5182d0a80e8"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-AR", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "134b7d45cf04b4ce4ac623f252789afd46c53b4006570700629a507f77677d93"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-DIFFUSION", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "1251b48b3428aad82fa97ffec4f1b91b88683afcc3c463344c5a1ffa807e280b"} +{"id": "modality-001096feefe6503bc4bd9020714599d1ab700f2d2359a5d33f94759251e00d73-ja-BOTH", "group_id": "diffusiondb-author-e24954c9756b04baf61ab299224fc1b9ceaebc4f6be4da5a6235249e9fc33a7a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b6dd603197fd66b1a6cfc3f65756cba1598a9c98cd93491ae051bc0a25c77e9f"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "9b5f92aad7db7d570f265d45a5692407102348e4f03ae0694e0e4f4f67362190"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "474e4aced9246d6ea9072c02b9342721cce61cd826b243cd4bb71f73f83935bc"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-en-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "6c071c9aa0f14165c478c39e9057f2c8c5402d92d10354d587f60e5e31677bb4"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2b97c264514be697bc808e271accbab5d3091717ec47e1231d8557808ce9977b"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0decb3c41851059ad1271e27646c3ef13e151d4cb8da936f433c474c2c3663c1"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "01d4010011a5e768c9357d6929d8b6bfdef97cd5af0fa5081308a44c20e87f3d"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "4c6f3ffe79ebc753059e1b55d916dbba0262fd1bc9455a6e795ddc64ed3aad36"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9eca3ef14681cdcf954b70e2bebd7bf641fbc38437efbb8ec350dc00c7f1ca6c"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-es-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "cb77432414c12da1b50bbc5a8833e07f56bc2609ef5ba02cb5035469179dd304"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4f8b03b15b1332fd3c4dd1db2a1eb3a0c4df6f1be313fed132274335125f54fa"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f7134803878b1f79948c15f5e5e6427c44ac95118da32b90618d16aabad12291"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-fr-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f270d2149664fa0a8d5607daadcd9221eb236eb6e7fc5d5e4dbf160a63291971"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e33bfc083d73584809b4ace1c4065e0e1a3e95969fd8aaef5269b46b796a9818"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "44a44698750d9a4673644aa641d35efc6a6d7a778c193c507972395a2632d3b3"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-de-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "18625d0ac7de018408eab8cc82ab175d32fc98dcd9c9c79330e5850e51155ae5"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-AR", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "070d274fdd1dd5f54bba1243a2f387d8dccb4cf10b281d765198816b65da7468"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-DIFFUSION", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "5de762b16d29c3c0a1ed75d865e81401e015a1ad071c80cb2509c51e58de9a63"} +{"id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-ja-BOTH", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "0499fc72422e32b2e4cf6ab1234f5004a9f60f362c4142bd8a2877c99a2d7a53"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "61e69a9e2a4447ae7aec9d5cca8fc14b5235b503716dd5cf607a850bfed21622"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "aec1fadb381347158f94f16112bf7c3039f144ea651e568ab676b4515d3a93b9"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-en-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "19ba76b3f3f6777d2f988f8dd24af64ba6f6a5d0ed473a35d2c2a88385fbaa97"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "98ae60d7495095b3e118bab213bcfbecd2754b0fa6128e49dfb2a9470d4c2432"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "16508649e1becb536dc404d1f018330fdec75cbcd3179a7d9458dbce53e9f67b"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-zh-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "2bc963d82eb9bb7726afbe1b66f1b4f8242b76e4ef754801d48d7722ac714467"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "58c9ffbbe84d3c40b31ff30cb394ad272138aa02eea5b0bad338205c0e3f5ca5"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f7dad4f25b80492386a8b1a86129532c758e78c3c7103fe0b859d92b7e766e33"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-es-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a2bf97fda530d9200a1eef83a86c3cd985b326e9d09ca20bc2bc8ced4d8cf615"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "f1f005deb80eec9b028f96d5bb7f72b75c760fee9528182a6a1ea58b06438b5e"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "722da9078452e171bf92e4603f3542b677284f54a83b02c8ea0a2417f83f89e8"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-fr-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3e55c64a2a198b1ea4dba380f70820f311cc55fb83fad5c58c933d4bab3cea3d"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "1a95f607f6eba625e825231384e637bc8382fcd1d20a8871a72da9864f7bba1e"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "571b3e53fbbb26d9aefcedc2ccecfff4283157a49e14701267303f02cee2561d"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-de-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "3fc4890cf9c63e278058a80331e10953b73fedc48616ca070f8312ed3ed502d4"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-AR", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "f433af777307bbac0e82529e0f8da7bc11115cf7a79ed2ba644754914fe6d331"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-DIFFUSION", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "b3e2475c9c5715cbf7ffef5d791047b55238bac537ef57fd4e6a8e2f2f9531c2"} +{"id": "modality-00125a2b729c30cf1c38daa0253c6f0370ed059bd925ead00a04d631b7f112c3-ja-BOTH", "group_id": "diffusiondb-author-5905d1cff6c5df7dc7c40bb4c9e43a1fbdf1ef3aa7156f075b4e1dab3ff82443", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "c9d8ca741f44df4bebd3c36cc2ea94ddacd01b9f30387ebb692296210ad8760f"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "d942b5cf6a4110c0d90cdbbabc18723ebfca07fe775945f85517d8aa3e5ac16c"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "348ed0bf881f2cfe218086551291b6c3be7fc4c290e3e02f8b74fca884fc4248"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-en-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "5270d3f72bdf2af2bb33aa6ad9f80d9cc99efba70e9ed9fb58d0910a6f8d0c19"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "97dfcc9c344859cc645636fce07e56114a1411aeb03b8b6ad78094753528f67b"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "1749ffa9677e412a49bc953770bf458b315ca7cf2af3b64b30371e05f219b1d0"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-zh-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "010b505cd043e0e10dba88058bcbfb3885f40ea84acd61a5ae64c8581bad5392"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "e358e74cae47a5acd589c0e01dd12c4f4b128893f1dd6f4f746403fe418e6a34"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "1523ef694dc0dda09dec339bc5a731800f83d5ec66f2f0fcb9af6932343089e7"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-es-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "aadd37ebfa6d5a28b65199eee737913d2eb3d3095de7aa0b2769803d21476d52"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "83fa2f87383dcaf3991bb656a3f063d2196344fe24ed5077c4311e9947a3fa29"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "fc64b477e98d0833858b8bbda667a355b23e6cda8d645da8f016c2445b938a88"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-fr-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "b8a8e429749d0942ac59ec9e768171d5eb28866b32e6021d250e81b19392f863"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "f92c8d6ccab19f7bbd5bc8cc2554cdb455c42b244bff4015460464711c388a9f"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "aefc74275793e5f4ed90c2bddcad544d9f26e0e29177de0aed9cda2bb0773205"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-de-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "5e8d38daa56ec59dde7cf387cbbe881335f35488b08101a6cb21094a3f7ae735"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-AR", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3a14b724268a50c03d43a9afbe65366399b340e8a99976e3791ca6cb9ac04937"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-DIFFUSION", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "6c8a421c7791e87b0d629876898353dff4c7adb2bb081b9b8afbecb5b0e7d0f9"} +{"id": "modality-0012c32e45e1f266a255b46f74ba6361e338fb22571d5182b92e3854fd319bda-ja-BOTH", "group_id": "diffusiondb-author-244d5169aa0646f6802dfe51b0f4b2af3565065171cf412fd890081c4f43d8dd", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "3fa26707760f08ec25b932d6dbfeb4631d537be5c0f5740904adf9a7fec52e43"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "bf17a83735ae676f8648a1259d5479170aefec8b17e79198f4e7435b58f32400"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "ece7ea67226434e19f17d8848faa2c09bbbc364d63873fc91f727ede898d383b"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-en-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "4e372ecdebb4fed302b9048f7b65c3e27d60a6b789402932032d968aaadbe808"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "3011068156b1fbf692c7042805c7a2d25ecffce342e241563a4c23b5be6cc6a8"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8313af56988c0b9e7a74685202ba9995c1589c5f8b3d59ebfef3a943df6d2735"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-zh-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "283e6fdc32dbfa5e6097dd2345854d107d98f8ac18e0adab0175ca46f14378e5"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "a3025edb08a3f90b94356af25599c749379e8051e5825ba7becc8b8b6fe9c9ef"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "9732a811f927e439a397f607bf0545eec0ed2113898ccf3c1655fbd7fc1761fa"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-es-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c3e10c2defe83bfbd768f361049798fb44df0bf062da716bc73252d14dc68b93"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "2e1e06032cc18e2140cce9beb3f2aad25655881ef253a0e193e36b23ceddef70"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "ada15de07e6810b86701a559ae50d4e409f0914faa1117e2b7636c5757a7a270"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-fr-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "33c2f48077ff287107ae0fac43f47fc66f2e60f123b8e5e16fe657283f4e9249"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "dadb91250a60d59ba28aa86c214b452026dccb1d1f9556814d6784a8510f7e26"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "ac82d44f6a2f19a6dd55ed4359a0f4a5b83c12caabbf2313cafc02c258812740"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-de-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0eaf0468ac59675013c9cb0e5b5581b2c3b0bd9c1cef799ddc89bb79875c427c"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-AR", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "7c88356c55332beb041cc0ea401d94d1c1765ad4fe3e0412e9366c0366f7e336"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-DIFFUSION", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "121d7708ef7b2de8d4abf50a33a5135250b5e3623382d6e9a399e409d803b1be"} +{"id": "modality-0012e8e7d096557676bdf4882e01c211fe69035324f3da484fff8c8eb6d42ac0-ja-BOTH", "group_id": "diffusiondb-author-6f6f0b5686cfd2db7b0bba44572a7a35348b3630a4eddbdaf37cfc135c00164c", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "33d92a5574bf8086809dec02f9b861360e25a00c8ba3d3bedde2c34d4a5f7208"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "a474d55654ec61c458aa45aea97da1e7853b6c05666232beb7b9971df73e58e1"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "a4bb3c681db0e657d63da684adaf7398a93b2b3ee6c14e215c6adb4ad5c326eb"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-en-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "9b3afe48ecdd365ca4794a41aed378e096dbd2167c48f43d830abc226ad8ebaf"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "a6e1d982854dde954aaa571580af4f0cff37458a46a03cd068cd025fe8308907"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "ca06a1e31a598930b5e5ee1e908a8020d19f6b168dcfab7f228657ea2c86c401"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-zh-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "8f64a40b9d06f632bb2ce8bee03421e874e4a3a8244bda225746cfd398b0da99"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "ea256aed936b2a6d8eac117b0cd4c25b9736cfa36b13d79c92c7e38b4b306742"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c5d87a0b469104d1db44779b4524629329f32193b8c7d55364f2e153ca7d0fa4"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-es-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "68482d23ea8d98f63eaf8e59ed8a986cfcb3efd1e735c120147973cc3baf0232"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "daecea81c44f62ad4c4ac92aab77a36fac527cd74c5704edeafa46bf6412e285"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "52fe2a66a0719ee219fbe975c19975e7ce457e72fef8ea4115d1969cf2f0e280"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-fr-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "deb4c9c4b767d0a34c7e7b9ef6c8607aa5b8f480c6756148bc8a0b9e48ce637b"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "ff30e584c47d4977587136745bed73bab878712f077341f2e7b2ca8e7c768d86"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "8455b1fdba2ff9701e7129a7e75fa13111949a39731f8a7cda418bc72b44293d"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-de-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "0329134308231039f42c8694d54bc51d1d19e3b8d6bbd7db4aa0b497ad02bfb2"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-AR", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9a02a5d13b19f848b7eeb697f807881be7964915a00736028fbc68576713df7c"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-DIFFUSION", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "085f95c07277a8edb740176cfa5ece8e2afc727bca5e2b4d767a3eaa2b972af7"} +{"id": "modality-0013f5fecb5f594653d634a6b98d2b02724cf0dc72e6b8d989e2bec4489ed2f5-ja-BOTH", "group_id": "diffusiondb-author-c06f0abd401fa4ebba3e3331b7ba45e310edca465c8097e774dc455d336abc0a", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "9096a2c30ec129ba066c22f6ea049a8a3c71bac9adb9ccfbad26785b65efbbd7"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "b6b87c3030df4c31f5457e79f35aa28d3b74a7b2ebcb35e3d11c6f26776e4468"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "40e93987c39d973cca1c91e96591cb7001fb5b606d7450baf0c9ff39445bf4a9"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-en-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "661721f867bba5dbd1adecd7ff354866c60ec074dfe13fe10b96f8d8147c48da"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "4b236c780d0cc4edd5f9742394a32ea82a535b8adc3bad7e2dd092541c332fad"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "64e469304e1ebf895eff410fbcc25f24694ea8ab17a6ae96edbc1e29a3ece38a"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-zh-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "571b73fb0acd19ea39b448277b313cea3cd012bc36deb8c30511f4de208833e8"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "c2ee3f25372a1eb4720ec7bc60bbbf371768ac2988f71f42be9d0c26ba3a8a97"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "d893b0e1b7fc88b98b74ee69be139a5e53ec34d5888b62c41a8211a99441957f"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-es-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f90efed6c3563e3ff7485cdd1e902c735d602c48c79ba9f2349a40794b00b346"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "514ad0e2a2283aa9a2b5c2b35e9945b66d15b4cab330b372ea21ebd463bab865"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "e9eda4a5124e68ce1c667f51f836f6c961d52c91772eb6fd81316ce4c1d58f72"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-fr-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "8de02076db0a8a8199c2023b44956ff31e3b6028b449c929f4ca6b8717efd323"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "dbdb96526d06a17b94559a3b5b3b11946900be1867a5055a4645897f9404164d"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "e8e2235d88ec8217db3f7b3c21a233b6026ea58b1af441be74a6d48e429f2f15"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-de-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "d2594cf1cab70e311c0363859840926dbdce452b3ce41bcc84f958cbaa83eca9"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-AR", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "ca8a24d61086505ce8109d08902d5aaa5a6e76da4c59fa5754c4512bef4b3e88"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-DIFFUSION", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "dfe3a4e93d33a5ac5a9c26aa45d0ac0d03e36aa0532c252a75916eca84ef915f"} +{"id": "modality-001417811481146cfe265f0deab22910e262d497f941da159f5fd5c7d48e65c3-ja-BOTH", "group_id": "diffusiondb-author-3247ec12b8a7dbfeca4964807467157ac5f0067eaced955b38ad5581e350558d", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "835bdac50e166da470cc6da0eef3524f6b0e23f5f3050c6ee11864fa2dbdd910"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "de9a13d4594a95f84279fa478151f8c18e25e44cb3d3c852be6b067ac7389124"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "3c8eaea8f5e525f9d1ca31f7ff7d782d8bb17a5593944a43fa90607ed6139189"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-en-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "en", "length_bucket": "short", "text_sha256": "a960ea15d19b4bd35564133da42afc31fc7a16eb12b3dbf4b5ae01443122927a"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "0efd60b1699093fe9112146334c88c4a73bb8af94a4c5a384fb1e43df6078df0"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "b02802dd021b64dbe64b7becbd6c6b962a966c4bf2a841aeddd5da4f4d9fdab3"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-zh-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "zh", "length_bucket": "short", "text_sha256": "d72252ff4b91f4b65e0bcd3c3b0a02ec1d0f3b60423e594c052f20fcc8c0b1bc"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "fd80d6bc454ce40b2576f1667d91344458fde3707e7c046253bdda3fcda79786"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "6b90748f741436b5bc52660ee3eee8b9ff709de630b924dfdb0e27255913d586"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-es-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "es", "length_bucket": "short", "text_sha256": "f548bc81d178ea0b7546bc6e0d16b02e996587c6b2a3654013e53f95eb180103"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "92cd9e4c284c10657dbc82c58dd6387b42cfcd46b729c547a1a9fcbd4e1767e8"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "3e5af00b0874d9fc52f9a6da72e9616253829c9b0ae28df9c5a51410030cd575"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-fr-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "fr", "length_bucket": "short", "text_sha256": "4d6106ef683040247fcc0e782ef43d71cd137e1faca60997507efc0a35acc0e7"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "87d576ed05af3dd28546ad8f46d33594eabbe791429a736c98d1c61ca743d79d"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "fde55a35871cbb02c2e0a90299f5138f76ab33a4f6b13e38074a4b0482095644"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-de-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "de", "length_bucket": "short", "text_sha256": "65ba529f7648b79bef1571510e68b06e85f631ed93a4c5553e62200dc912d2ae"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-AR", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "AR", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "d88374705b9835705889870fed22232c28e961eec0c0a32adbc3c383d1c6976e"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-DIFFUSION", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "DIFFUSION", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "fee648e068d932c94f271daf7d75ea9cb7a35c7ea109d845167d09b3f52a2644"} +{"id": "modality-0014adec6cd62c1fc554940dae02433561070c4feccb9bda25952926a5151ec1-ja-BOTH", "group_id": "diffusiondb-author-5f49f23c11d543410887a5ef057070d9e1a7d9b74fcfdd4c0a026945d11276c4", "label": "BOTH", "source": "authored-output-contract-with-gallery-caption", "language": "ja", "length_bucket": "short", "text_sha256": "28e33512788592b9b1d2d23bd385ff846d93d0f050d8dd1a8ddee84053257220"} +{"id": "modality-text-global-zh-philosophy/test/165", "group_id": "global-mmlu:1bdfb9bc63239e7b3e78a3f4cffc04cb20b94fe52c6a06822faabe565f77c3cc", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "7495d781688ffefaf4e9da5d48df599c9225bdb937fa6618b6f059d6bd288569"} +{"id": "modality-text-global-es-human_aging/test/127", "group_id": "global-mmlu:fa15e3457e6d2376a57ae84f26daf8ff5da762d7ac8b54fce5e30261012d05b1", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "eb9e6ab2f286c47c81a28421a61748dda4937169168738de622299ac9d5a12a1"} +{"id": "modality-text-global-zh-elementary_mathematics/test/289", "group_id": "global-mmlu:e097f3597e7c6d6f9ec611fdf2ac2170944ec7c5eca8598fa8ea7f40764fc8dc", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "c5868af67c36f61987fc83132c6506641952fc2e2d491dc526d009fc3d706b08"} +{"id": "modality-text-global-ja-professional_law/test/570", "group_id": "global-mmlu:939144aed9e43db91ccfb92bae5c975f58af34c9864940c7e5bc35d31b9eb6e0", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "8b82bd897f38c4bee8ed6e7db0b8c7969a68da80d1c312730c77ec7b32be4fb7"} +{"id": "modality-text-global-fr-professional_psychology/test/359", "group_id": "global-mmlu:4f3f3f158c141fbfa8d1655f4c2116fc3657f10795cb233b51a9e91b12e6aa22", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "661fc679363aacf2a39a4995471acae548dcb293ff9db34563a85b17cee7ba2c"} +{"id": "modality-text-global-fr-logical_fallacies/test/136", "group_id": "global-mmlu:1e81c8b94cf5cc49f9761bb3349cac43293e8d565ae1f57c19c6bb12acd95809", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "7bac4b993be14727a0540496079b7a58c19f1aa9cc503a39e96a0c08897b5e97"} +{"id": "modality-text-global-de-moral_disputes/test/116", "group_id": "global-mmlu:60795bd96230f2fc9b3f60ad46df42c2f4b456d378485448a85db76300f39ea9", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "83e14410df4112c8d77d023f5fc702fdbdbe61d4a9c147a05e1c4364466041f0"} +{"id": "modality-text-global-ja-high_school_physics/test/67", "group_id": "global-mmlu:dcc900816935402b0ec44a18d08ea268c3b1f5f255372265d0d337956376ff19", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "09e4791d481e2faa302b4367781910d681a59be116fd51fed1fc1a32ee6e6ed8"} +{"id": "modality-text-global-es-high_school_microeconomics/test/172", "group_id": "global-mmlu:3591b4399e98434173dc0cb40a19b9c0b7b662ec8befd4c953827b45e06887db", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "e73ee8ff3ccefb3c7934a2b101308b28cbaaa22a717166cad8016d7b25872543"} +{"id": "modality-text-global-fr-high_school_microeconomics/test/123", "group_id": "global-mmlu:79441f43ba3734b20a3d72240675ca93c56dcf492e9534daa05c7f53af9bc808", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "055f6a409f8ab9f5f5e27b424b6b0815fc56f44a2ed92aaa2d84eea5641e2396"} +{"id": "modality-text-global-es-professional_law/test/986", "group_id": "global-mmlu:c2feb6df6587fe347ca7091950a69bbaf7de2d62c98d9946134eb4efc430feb7", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "d8e067447526dd3eeecc5862032b1feb11656819399190b39113bc8acf521266"} +{"id": "modality-text-global-es-prehistory/test/47", "group_id": "global-mmlu:1c248b45b4e09a04a904459af72d71ecd7d9eb9a1d0d4b8caf68c93f7404ff0a", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "58c884cd5e44e526420ceb712347f24d54db5ed9212f2bc71d0feec9b81ac3c9"} +{"id": "modality-text-global-zh-professional_medicine/test/56", "group_id": "global-mmlu:1fb5e62e37558035dc78ebad507899e403ba83e5f4a3915d0c542de6c8af39c5", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "1efc164c770d5fef2c5a80e1be6ade9e1355b48daa1c60156c307ffb97c50fbf"} +{"id": "modality-text-global-es-miscellaneous/test/639", "group_id": "global-mmlu:82d085754a71fc3ec400da155195309130498bec26064444c3d485b31b4c36c2", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "bfe47d52818cf7e6c92c899241c1678b8aff09fd0f64993049ed3a172eba14fb"} +{"id": "modality-text-global-es-business_ethics/test/42", "group_id": "global-mmlu:2e6112886c42ca3bc346d61b29911b2f1566fcf990261125efd30a727d8f46d3", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "d477b7afb7ecc90de6ce2a3f5a5eab93a77adb17404525e9c7fc108214dfef63"} +{"id": "modality-text-global-en-virology/test/107", "group_id": "global-mmlu:bef3d7084f33ecf5bf551857d702c72b8a292cc5726324f7a27030f11d91f05f", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "290a5e64fb6ac7bf20b461706760e935c569add77c6345ec4e54f56745610b9b"} +{"id": "modality-text-global-en-moral_scenarios/test/304", "group_id": "global-mmlu:c919ee9e04d754d3a4a728bd4a62c7b15e7d1e7804ab11c65bb68f103efe1529", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "14431b731a3a36d6204528df00b1957635857ae007d9f0fd161625a15de2307f"} +{"id": "modality-text-global-es-high_school_microeconomics/test/157", "group_id": "global-mmlu:423476a5a2d34d9921ef2772f681057fd6232038c5f8a0081824eda410e8ed85", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "7b01acb956e6e4c852784b6e6b5769686ac042f6e7a1fe5dfada63de342e58b7"} +{"id": "modality-text-global-ja-moral_scenarios/test/453", "group_id": "global-mmlu:bc14db651f205e37e7fa9a94c0bda2488582bb34d355221994acf1e71ae0c585", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "5aa78b0c732237f5f04fa3c804ed593ac76e1c99eaa8ef7fffe6ea34b4707b30"} +{"id": "modality-text-global-de-high_school_statistics/test/185", "group_id": "global-mmlu:6105cd476177995fa5d3099cfe7891c2141fa9210fe189dd0f9395fe26df661b", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "a47c151f700dcc68d2ea8d54c30dc98d171fd8d16c9a5d485a5db4bd1e2fcd24"} +{"id": "modality-text-global-ja-moral_scenarios/test/73", "group_id": "global-mmlu:685e3433ed12bf65b9fb0b65d4905bf79d0bf86a746ed2fd1521b337c3ba11e4", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "9c2b72fbfbe3ab7f7f7693c137e805e2b441817d15be4260d0a7ff73f38804c7"} +{"id": "modality-text-global-fr-high_school_macroeconomics/test/270", "group_id": "global-mmlu:a5ce657a787195832456c6787491db371ee94bdc2e2ee343fd4a516816706197", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "5ccdb17a15f10055edc6dd56acfcae6b4ef866682244fb897239244741e09936"} +{"id": "modality-text-global-en-professional_law/test/539", "group_id": "global-mmlu:274f4c9e85a73914eb0a66232c18509206e48adffe92c73fb3db82465b1dc3cb", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "6d8b35fa2d00f0c7da75c796fe6871c65b3ae71365c311ddb2a3a6ce70362139"} +{"id": "modality-text-global-de-high_school_macroeconomics/test/172", "group_id": "global-mmlu:f773a2227d3cdd3638422defafdb71daae230f7b7b981ac1738c307f2260557d", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "ecc8de2f3d026f54b66f2256b839f56c72a5feb25f3e588d7a18c36076957d2a"} +{"id": "modality-text-global-en-professional_law/test/9", "group_id": "global-mmlu:0d6e4d9573d9bf663cf8627e68c379d98b54256ffac6e7c4de39c648f7de9601", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "684315195018f817e0bedcf3a732740427238e82e5a42e814ceaab3a7ea60b9f"} +{"id": "modality-text-global-fr-us_foreign_policy/test/28", "group_id": "global-mmlu:9dd332ecde8ab3f7deca2b3a46c72fc701b885f9273a1c844fcca24ed60efc70", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "79525d8a249865159af459de373f2d12f916dab43c55515ce379352de18df355"} +{"id": "modality-text-global-de-virology/test/107", "group_id": "global-mmlu:bef3d7084f33ecf5bf551857d702c72b8a292cc5726324f7a27030f11d91f05f", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "6f8ed00377e39bc7ce5c04f078a8b6a2b60ef5b69cad5f3e4830110442f6fb7b"} +{"id": "modality-text-global-ja-moral_disputes/test/331", "group_id": "global-mmlu:4984eebc8f7bfe3e8a736ba9aa152b01f6187c175dce19066ccb272ddf66e2bd", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "378f62747729e3050bdd0733e00cdd6cddf6fc6e16d7e404891b08fe916406f5"} +{"id": "modality-text-global-es-college_chemistry/test/55", "group_id": "global-mmlu:80af83950f4a97fd6650868256a237773ff6e5b5403183e2e2653697cd158d59", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "85eb19a74595989a18cc5cf0b98a66728ace7d87b9b72723e20f366d8a0cc601"} +{"id": "modality-text-global-fr-medical_genetics/test/38", "group_id": "global-mmlu:215560ce1bc4a30f4d234ca52b459397a1ce9b3124a76736b48c3c5ede20f147", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "3ed7f0c0f7fbea7894d031c17ffd7e03b53be1cc545e65f3ada2b98b4dde572c"} +{"id": "modality-text-global-de-moral_scenarios/test/833", "group_id": "global-mmlu:4a8d70a106d43215c1f712c6ae2bc45a3935ab5406bd28b1feb19105083268cf", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "1d0889db2d5a29007ae60851a16e203e4f88f45c8bd567b296c797e90107b03b"} +{"id": "modality-text-global-fr-college_biology/test/28", "group_id": "global-mmlu:586d60dde19d62d66685dcc190ea147e7fcf2e61eafa3eb3919df6a75d5dc324", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "d149abf1f21d87cf1a35621ddf16d8e358f34e7e8ba6198fc31473a2844b56e8"} +{"id": "modality-text-global-de-world_religions/test/23", "group_id": "global-mmlu:8c35e5fe516772240e4d3144b964767de477ceb90f29e06c5448fe166b1a4343", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "4916f2b490009573ef59121acdd0fb886519a457d756252fba465c51eef9770e"} +{"id": "modality-text-global-fr-international_law/test/92", "group_id": "global-mmlu:03c8bc8e61e0943680fde2f95f8077f1d670de03bf9151517980e1476edc2c6c", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "235e5ec3aabfa8cbf8359ab6438e60c922335a602e644e80ce88b89eb8f9c23a"} +{"id": "modality-text-global-en-philosophy/test/52", "group_id": "global-mmlu:928b960ae60dd6a09e03eac964e108206382bc677686c3db2e6065870e70c8e9", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "750585e7305330705e7e5722166f726e642a1dce1d84aadcd41234966beb8845"} +{"id": "modality-text-global-zh-high_school_world_history/test/209", "group_id": "global-mmlu:d60807dff139630330c827030fb305eaa608e561e4413acec112f383e44bcd4e", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "ae284702e72f939ce31d4dd215b5e08b3f46f3c8f205857a561c5425da794899"} +{"id": "modality-text-global-fr-philosophy/test/165", "group_id": "global-mmlu:1bdfb9bc63239e7b3e78a3f4cffc04cb20b94fe52c6a06822faabe565f77c3cc", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "20a4769f960c4ded89d145a33cc852cd0cd889738b2dd33fa3065222c2127979"} +{"id": "modality-text-global-zh-moral_disputes/test/36", "group_id": "global-mmlu:b76c4e8d1b3d6544c8e50a55f9f93fa7fa6e28284e10723f86e0d211bb762bfb", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "18d3a7a4535eaeaf2c226468eba36ec344734e8bbf374b511f221087120f9005"} +{"id": "modality-text-global-ja-high_school_physics/test/133", "group_id": "global-mmlu:03a50815acbaf87088cb18407395d81409ef10d6a40943c50c6a080eb419145e", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "ade844ba119355359595040b2d3d4c78842c8b5a88895fcbfbe583b2fca224b1"} +{"id": "modality-text-global-de-high_school_statistics/test/145", "group_id": "global-mmlu:957438a0f23bb34e584223ae611378ac46669aa13835071e7a04d2d7adba381a", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "6cc9d05a0d7f464581f1e5b163b6ee65993e192b9ce09ffecb91ba28ce073f6c"} +{"id": "modality-text-global-de-high_school_psychology/test/436", "group_id": "global-mmlu:6ecf766577665a3bd01b41a156eea1159f76f6badf9c90d9beaaf5e5edbdea74", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "2c8fc56dff4be5ed4e4ce306b917a64b18c65f6fbc47861af8e0794183a32219"} +{"id": "modality-text-global-es-moral_scenarios/test/833", "group_id": "global-mmlu:4a8d70a106d43215c1f712c6ae2bc45a3935ab5406bd28b1feb19105083268cf", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "6c9e4ce39484121396298554c8535a4b96aa3925e3c83a99764e72dc1aa0520f"} +{"id": "modality-text-global-en-security_studies/test/107", "group_id": "global-mmlu:8a9f683de2a72f387ae1717702fac6ae9b6a85f6d2d04ae640c9a445d40e2a1c", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "82b173d6f14345a1910abd3b53f2604ea2b3de415e5e001f8f1441c818b4f1bc"} +{"id": "modality-text-global-ja-moral_scenarios/test/848", "group_id": "global-mmlu:04aa4229b3431d0697756f2fb450a1755a7760a1f249456595ea40ff171f1912", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "227988255920a110561ff8528fc62615fb8e5371ffd29f5392e58ac2673508d5"} +{"id": "modality-text-global-de-high_school_psychology/test/137", "group_id": "global-mmlu:dd1846dabf30adecab585c9037fb315debb9085ec618c4d8b5230f07f11cbca1", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "12ec4e9138e988b4afea3c2a44031ca15b95b68edd15342afde79b6d77a36c74"} +{"id": "modality-text-global-zh-professional_law/test/789", "group_id": "global-mmlu:7bf905098b39d64fb64789728ff2cdb9dec8b198b7d87e5e7ded21fb33b355f2", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "bd03dec9157877e485886e2275c4d26735c7927046ef9a577f553f3fb1768c12"} +{"id": "modality-text-global-en-nutrition/test/129", "group_id": "global-mmlu:8e1c5dc3668f3f93e14b09add02da06dab3b0607c2bbdcd0823a948c1a73eab0", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "07db19f75c10e1e14216e05842b658b79658cbc81c18e870f2aa5fa16990ece2"} +{"id": "modality-text-global-fr-professional_law/test/378", "group_id": "global-mmlu:430bc7beddabcd93cc5c94ee66957ec79fceda294cdfc1603b6ec0ad08eb1cf9", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "2637ee86434cc768bbf33669918d7a11ad234d3810fed5fd3d61877810f9f1ff"} +{"id": "modality-text-global-ja-miscellaneous/test/728", "group_id": "global-mmlu:ae38faf9551958ac76c409cdf38b07100522e27f8fc1e39620ca9719862a9245", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "cfad660d1b46e56186e4c69cb72bc82b3339e52462a4ca88b22234d5fce9aa4b"} +{"id": "modality-text-global-de-college_mathematics/test/53", "group_id": "global-mmlu:213133e2701c38ab7f024cd3d8205a7597b52ec2535318ee969e554884aa5bc4", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "435ea4416f1093ddf1007c8544c95ed7db595a9a075304acdc2d7c9262b4440d"} +{"id": "modality-text-global-de-jurisprudence/test/41", "group_id": "global-mmlu:53583ca05773dbd55dc7b9bd22f4231297fd6b8aa789037c6dfefb21d9708c3a", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "49581ff43c6947fa7c96172f3f37dd165e0f1decd825b3a0a2cd8e08fd851893"} +{"id": "modality-text-global-zh-logical_fallacies/test/152", "group_id": "global-mmlu:07a759349433b8821b07ce5a62bd10c58f69fa4b52a75032466d73015bba2163", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "bee992254307c3e65b2c456a7a8ca8fd04c4ec930b4f803c59de0dc4002b52c5"} +{"id": "modality-text-global-en-moral_scenarios/test/56", "group_id": "global-mmlu:aeb56347cd0d8846bc4f48de079a8e9599126856b06e310a5731c5adebccb908", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "4e4accfb464f785bde4a88f4478ac9e33f18fbdeadec1a0a0325b8063a8e5e34"} +{"id": "modality-text-global-zh-professional_law/test/294", "group_id": "global-mmlu:820a8cf2a639174ce9127e2436b291d62de279e5dd9934401202b1d48da8c76b", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "1d35f609fb6710804a506d7b4d66f1fb036c2bf9f06ab85706c42aab701890d9"} +{"id": "modality-text-global-ja-security_studies/test/19", "group_id": "global-mmlu:455b1d5b3e69498ed059edf0edfc07870da12428bcb61ab3a35d858502fbde44", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "e3f5ea40635be39625e57412ec7b4b7c78a36395e26da098c5b1a50d243c7edc"} +{"id": "modality-text-global-de-nutrition/test/301", "group_id": "global-mmlu:7e9689610770055dfac707b537c5b44a6a4451f62074fd7e599dc2dd3e5f3d5c", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "cc85a7c99d609c5d37a0821350fca15bd5caec992e5b7fbda49f5f555836397e"} +{"id": "modality-text-global-fr-marketing/test/84", "group_id": "global-mmlu:c09daca7ab3da240fb263379162c93a4b8efe125ba966b936eafcabea10cf757", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "51f1a14b3bfe39021b8fabfc49333cd454ebb87cd2242f21ff344ebf8d771e82"} +{"id": "modality-text-global-ja-professional_law/test/829", "group_id": "global-mmlu:5e08abb55e2920b7fcfbedd807f91438b7b749e394233788bfc7da0e4f2db53b", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "d8d29fb4fad4b906e6863a6c22bdee6e1e8b42e8c887bad39029850f5821bcd4"} +{"id": "modality-text-global-es-moral_scenarios/test/51", "group_id": "global-mmlu:9a1613333a25d95d54a616169d89cccce540d6190b3aa6f232c94738fe2aea5c", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "c92890e8bf7330ba45063d33e9cadeb3c40c3978e09291ca26002705397fe17f"} +{"id": "modality-text-global-es-virology/test/64", "group_id": "global-mmlu:e37850319254de63244192cfe23dc5298ca55ec47c31bfed5588746a42ff879d", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "c9ae23dd875fddb2733308e45dd434ce03ff83b9fa78060257f9e5ece69a453c"} +{"id": "modality-text-global-es-human_sexuality/test/1", "group_id": "global-mmlu:9dc5388e3b4084d7854f53b709e1696f96a498cfa2713dbdaf3f969f81c1a7e2", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "86e33d9621a29d528f27ba86530373a7dfe1064876acc0e54e020e45334e4730"} +{"id": "modality-text-global-de-virology/test/119", "group_id": "global-mmlu:debc3dde5e20338f47ed7dfae45044d05f5345e2a907f621c85aad767b148b43", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "96200a57ea969985990c472f1c7c4948d6dde3532f62cf9adbacb2f637d2a17f"} +{"id": "modality-text-global-de-miscellaneous/test/108", "group_id": "global-mmlu:5a64fc6d0817c3c1979495118ba6d3a0880d5a374a9ae2e1d6c6b8d3ebb6f8f4", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "7de7f7607b9b6a08e754eca5a73bf1b52f7a831ab94e486b326336fe5d3432da"} +{"id": "modality-text-global-zh-moral_scenarios/test/628", "group_id": "global-mmlu:7795701f590847b5121009caed9221ff03022f424510a5d6976d5964401162c7", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "1a5f903c7d27800bfcbfb93b930dfd36b13611f8ed98e6e7f4d1062586c13a71"} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/24", "group_id": "global-mmlu:40254a2692d9b7564819943936720c982d1a4b3972dac680232f645f3097718e", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "8a2c44c77335c2ced60963e3f73ccf91ab47d606ca2c5ad781485985be57cdde"} +{"id": "modality-text-global-zh-high_school_psychology/test/137", "group_id": "global-mmlu:dd1846dabf30adecab585c9037fb315debb9085ec618c4d8b5230f07f11cbca1", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "ab9bb91788d141eda9ebc94acd59084d590c523b6e5af4bcc5ca9d854462cde6"} +{"id": "modality-text-global-de-professional_law/test/1383", "group_id": "global-mmlu:7eb9467697c98859e76e4618ed3b7b4b4dbe306aa4c6944928e44ed90ae1f934", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "2653260fa8cd25ba57ae819551dcdb7a92cc9c2b7b9dc49711d1239c43619099"} +{"id": "modality-text-global-zh-international_law/test/17", "group_id": "global-mmlu:4a793d240bf69961859b5250d12e1c0215bba7ae239c7dfe6a413ea32e3bcb63", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "d6c15f5de26c6981bbda854bd0a606f814ce66741604fe136e3a1237e592d78e"} +{"id": "modality-text-global-ja-marketing/test/160", "group_id": "global-mmlu:507d00d5d5bcc6e7a52e686093f12eb373e3775fef1fbb0085e33301987355ae", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "3a09b524984e14a03f61fcc6eb6b98b5fc45a8a84ab431a511ac85b789aaa61d"} +{"id": "modality-text-global-es-high_school_psychology/test/250", "group_id": "global-mmlu:e65ffaa32473c3063d19d9bdd03e982523055bc0ba1b4260552ed278f40edab5", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "47ceba4f7eb0f0abb72b1d1c2c49952c76fafcc11691ef3a3f325868009580f9"} +{"id": "modality-text-global-de-computer_security/test/86", "group_id": "global-mmlu:fcc60d46c92e9e04e190740ebdd48043b10ead829939e1d0688ac4a0cb5e2457", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "ec91c0fbfd60acdbe3de1edb3a630f57897cfec28e697c810ad791fd1e7f248f"} +{"id": "modality-text-global-en-miscellaneous/test/251", "group_id": "global-mmlu:6227855e642226002bce71a7cc7deb008f28dd1c720515eed01e00d77fab8c7e", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "efe0d6b68e46d5949d1ea0c58d6a8af07ea4c9b64abcf4d08ceb4977fd7e7c2c"} +{"id": "modality-text-global-en-moral_scenarios/test/477", "group_id": "global-mmlu:baf51bee64cdd219096a6df7b7c1946922cf170f0f29e24a4190c7eceab76dd4", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "d918c2a033826add068563059037978cb74dcbdadd773c399691c859529d952e"} +{"id": "modality-text-global-fr-moral_disputes/test/224", "group_id": "global-mmlu:ef25504b824d64e70375bc0c42a9d5494b3412f35ec3e90b62679b42932ab484", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "649a4f32b39a2da68b9161fe229f493dfdba4dd482d0ca53cad25b36d55b4597"} +{"id": "modality-text-global-zh-professional_accounting/test/54", "group_id": "global-mmlu:5f2623d45e88d8d843426d8b7a5cde9f3d760e3ffdff94692fb3eaab9b407429", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "2cb096fda5de9ae0694852f526569e83c0be32ce8502baddfcaf87d9d15aa0d3"} +{"id": "modality-text-global-fr-professional_law/test/1020", "group_id": "global-mmlu:9d2e18865a02ac19dfb3642911a6913a0d76da04e325d9c5618d7d473132936f", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "d1b64c18805a5ed1583aeb7c401f354dfc57044e994e06cf7dfaa19782cdf8f0"} +{"id": "modality-text-global-en-high_school_chemistry/test/88", "group_id": "global-mmlu:8f55b019a039f29fe6a3e0f7d71909fa61c8fdabeda990f5a4bf4de1d841e1a1", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "b11f5557981b5e2f9cb283ce187de8e52aae2d390df8612afcca78fc13a0f362"} +{"id": "modality-text-global-fr-moral_scenarios/test/94", "group_id": "global-mmlu:92bb12d7a8b55fade4f83aa7c30d9be286759f972f61214006d200844aecca35", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "21e5a282b10af99b4c1c50fc9ce0e9369e7a49e6be9b2e592dd9164ed98e91a2"} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/191", "group_id": "global-mmlu:5d2baea2c56aa5a5eb89131cc53f1c4473660f883d432ef4a771ed44f3ff60e0", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "496d8c188ae539ea49bf1445499e8043c243099c74f16368f6db535211e6161b"} +{"id": "modality-text-global-de-high_school_psychology/test/320", "group_id": "global-mmlu:50a34602ecb88b1d39041bd69bffe8efaf736de1c30a734da81418f658f7ddc5", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "542ac5095e3a13fb41d1b8ad0e6b25c51a7e2132176983e57e6cb158e88ed752"} +{"id": "modality-text-global-en-elementary_mathematics/test/315", "group_id": "global-mmlu:2843b5698480eda5a5bff225aa06010560923b38af89780aecd0c8b61bc4fc48", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "f575385f07dc1120be806fbfadb615a831ff6b05717261baadf0d50438330f46"} +{"id": "modality-text-global-de-virology/test/118", "group_id": "global-mmlu:383d051ba5259b6cf00f29ac90b3bfa1ca5a41ff3b53eb89c0be1d9160d99386", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "0778ac1c5d391725f8a9ce7536d1ed1140f1807a6c26ce460700a878201a6b02"} +{"id": "modality-text-global-fr-nutrition/test/92", "group_id": "global-mmlu:d4df9a3789d217ad538946000d5d4a69096890495f96cde74270aba9b816113b", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "1065aa6fde279815b6c23e3f0664e6465c67128bd193c942e9b4d7883ef1ee51"} +{"id": "modality-text-global-zh-computer_security/test/41", "group_id": "global-mmlu:3392f46c85b9161d666c1e34b9f7ef3715ab7005f3552a86013ca49888ba897c", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "e0982d9b747eb0c01dd9c826f859694ad4e7cdb9ca100f2778e629c3ca7e7e3e"} +{"id": "modality-text-global-en-professional_law/test/1072", "group_id": "global-mmlu:212b46dde9581f3c430188bcc31fc496fc987667853b8b337b6c7c143b321287", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "9ad9303f07c00e208a99fb0e3917fcc4528c0fa93e7b5ff0d852bbbcf50c9802"} +{"id": "modality-text-global-fr-business_ethics/test/56", "group_id": "global-mmlu:60b9127d10d6b0bc1e2801ccd837175d3429506cd3ea5e1f507a83a90aec0408", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "464b8ae0f507877edcfc15881ca7d6f23af92af96100fa871e96faefa46c734b"} +{"id": "modality-text-global-en-sociology/test/129", "group_id": "global-mmlu:853afffcbc0d43fdb96b08cbcf9b2411805a34332c5876f88119ec8fdc1b15ac", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "4cb302db66ead88eea2f13d77acade628ccd7eed4c17c72765390e4f1cae99f1"} +{"id": "modality-text-global-de-miscellaneous/test/413", "group_id": "global-mmlu:274eab02d974a0692bcdf67a00544ad44bd697b840f6c9e974d2c1e1f95ae0c2", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "a322f0ddbba975995b3d2f3debf11e383af36a91cf02e1755df37512ffb73d5b"} +{"id": "modality-text-global-es-world_religions/test/37", "group_id": "global-mmlu:e74294278473c72fb9e71fe663bde6e49268f9bf0176542ccb4d04cdc29bd508", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "c965deb303e989fbe4962bfdf2e3cbe97d355f1bb93ef0ac25f76d387c257652"} +{"id": "modality-text-global-en-public_relations/test/31", "group_id": "global-mmlu:7792dfcd8dfe51b1fb8736097851e3be2b354cd4eda218167355c2182efcf6e6", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "a922c460aeedd79b403e9692a22063701fabd3cf2f4b097a7b882212267c69d5"} +{"id": "modality-text-global-fr-philosophy/test/6", "group_id": "global-mmlu:0816e61ab008b9b5ce7e4d4d3ddffc2b175d6ae14bd82c0a3d19218b35bff685", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "84b264f5d4c9f7ac82831aff4ba7c44c0744efd0bc25bba7450f45a3fa813490"} +{"id": "modality-text-global-en-moral_scenarios/test/630", "group_id": "global-mmlu:4031731d9ae0c9f7aaa7978a19266f9487c08d318ccd85af1d9f507bf4529234", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "9be273c820d128fb8f0779a48bb947e6f1f8d8665b5adc29ed50e46494561db2"} +{"id": "modality-text-global-es-high_school_government_and_politics/test/135", "group_id": "global-mmlu:2b8e3bad76f42c885cf775d157b7ff98173528c08c1833c728e6225bf55c1672", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "8a90ca85b51ad4e6f7a7958e34a94bbdaed300941884af22a53c665c3948ed3a"} +{"id": "modality-text-global-zh-moral_scenarios/test/240", "group_id": "global-mmlu:8e30749a491e581ef570566203900b3a26735a1320cf526fde7d5ef2e8d1f017", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "a4cf6d701c34245511652efd1e8e8ee2dc4c1e040cd10640f8dc4fb90e9930b1"} +{"id": "modality-text-global-fr-us_foreign_policy/test/38", "group_id": "global-mmlu:896484cec6582cb09cf095e55008f4df4c6415265831b6944b58ace23d0a346f", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "7abd090451211443ec0f56bc095b7025505678e955054117d2948e8f35196832"} +{"id": "modality-text-global-zh-professional_psychology/test/412", "group_id": "global-mmlu:7a7745ef37fbc43c0f2ba84aa6495c91d6d952c9d88f382928539fce612a4a05", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "828d29a8d47bafb173ef65c79b272906ec2ad9a75ad3527c8991be9a6066488b"} +{"id": "modality-text-global-fr-moral_scenarios/test/843", "group_id": "global-mmlu:27071e830e93b4be3af49630485f83507005c8ad7d20a950ac016a17e02f0813", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "2e31aa4b12a3fffbeeb4fe62d4ff37f83d59e6a31e0258f019b6205d472c96ce"} +{"id": "modality-text-global-ja-professional_law/test/1248", "group_id": "global-mmlu:483659015a06af073bed6c7ac84312813e99f0c97d5679af80b8d8233acd14c1", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "ad8e4a8175ce86cce0ff36a7a3b678112f344aff792377731cb31dbab5751470"} +{"id": "modality-text-global-en-high_school_geography/test/44", "group_id": "global-mmlu:3ad287995b2c8fce1b442dee90d8bb831789e243b8034bea4935f60b0724c9b0", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "1e872326f2cb36b5c30d8e8f1225c2598120638ea5f42cfdf0dfe26c1bb990b5"} +{"id": "modality-text-global-fr-anatomy/test/115", "group_id": "global-mmlu:e517f919bf40d12c073011d5a9a467f2450685d96554101fa033e6573ec39086", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "761ab9a6300d32cad767fdecd064ebe968625ee409eaed5486df86fe9a3922cf"} +{"id": "modality-text-global-fr-security_studies/test/238", "group_id": "global-mmlu:6a161111dc48a96f442f375d7c606ba8916a0da1de2664f6006e57e0740b1579", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "825322abb0cee279ecf26a31fb56588d63af3a03157945dd96ece17c4fc3d154"} +{"id": "modality-text-global-zh-high_school_psychology/test/320", "group_id": "global-mmlu:50a34602ecb88b1d39041bd69bffe8efaf736de1c30a734da81418f658f7ddc5", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "1b7cbe40be173b16f55b2ea644b31154d0e2527a5a3b3e92feae2cb748179c0b"} +{"id": "modality-text-global-zh-moral_scenarios/test/573", "group_id": "global-mmlu:ef6007f487906b5c50da1cbbbac6ca10b6855be6c78367f00112826fd162d75f", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "02a0fde4cf1c27a54223a557d995054ffc25811f311d43df1b25faed0b726873"} +{"id": "modality-text-global-ja-clinical_knowledge/test/152", "group_id": "global-mmlu:251453c8f8e32a25491187a38798c920dc9d9c97d75c316ec897e734b92bad6c", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "0ce30e1c552dcb3a76cc3b41e2398d5e4d8639b323d2934a25c949639f17d8c0"} +{"id": "modality-text-global-es-world_religions/test/133", "group_id": "global-mmlu:c322ec70058f2bbf0bd33ac4add36389479bd304246dc52e8f159c29318fc32f", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "5a79ba86ad93f14d3f455af1b1cf9dce8730c0ef8c5558b014f50dafb7ea4eac"} +{"id": "modality-text-global-de-professional_law/test/479", "group_id": "global-mmlu:eadf7df64aaf49861d592493e0dbb37eb4609cd3419649a4082cadc64efa220e", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "4aefd3496a06a34ba65427118a60264a385979d6b38d261ce8e8dbb1e8b41e20"} +{"id": "modality-text-global-en-high_school_chemistry/test/21", "group_id": "global-mmlu:119011d8cad0b030640cb62bfdc5d813a9d47989ab91fa60116b3fe5071e0c20", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "62327fc13881fea5c9db663a49e64cace580c2c1c73fad1ff2cae6ee3dd82d80"} +{"id": "modality-text-global-fr-nutrition/test/145", "group_id": "global-mmlu:8853f2c9d2e13e7db21c56a7d37b7c09268be46d9d22fab0d083ef0e97a1e898", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "392f6ceeed79090e903c329bd4cfee684bce0037f91328f5170d095152c99cd3"} +{"id": "modality-text-global-fr-human_sexuality/test/46", "group_id": "global-mmlu:1b23b393d00f5112e0a489ad538706b869382f17f328d76fea19aabf40099ad8", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "4fb2ada0374b6fc8de5dc95f51a1ffca3ee3622aa90492cc9279ce0e37bf580a"} +{"id": "modality-text-global-fr-human_aging/test/29", "group_id": "global-mmlu:2fcc8135b7961ff758150bd72fe68d79cb242573ddc0376c2fcb77dc35aee41a", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "06097e76a34e09cf71acd2cd4964ac9a3d5efb4633e297dbf46fc585ac4ca7f0"} +{"id": "modality-text-global-en-professional_law/test/479", "group_id": "global-mmlu:eadf7df64aaf49861d592493e0dbb37eb4609cd3419649a4082cadc64efa220e", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "c2ca2f616398897f6f774dcc37fbdb74582cb297fc84a8f74811f8eca0d2281e"} +{"id": "modality-text-global-es-high_school_microeconomics/test/123", "group_id": "global-mmlu:79441f43ba3734b20a3d72240675ca93c56dcf492e9534daa05c7f53af9bc808", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "6bf10776edd36fe9a5df0c15854950a727e00f4bbd36ba1912fd35d00f8e6cb2"} +{"id": "modality-text-global-ja-miscellaneous/test/624", "group_id": "global-mmlu:177afd0982d9c07f963561e65cfbecdf03b707ce4574f23e37e993bd2d53ebc4", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "a0ffe69b1ce627615c9792b8199f36e64a5af6bee83cd34fce615f1a0046061c"} +{"id": "modality-text-global-fr-moral_scenarios/test/272", "group_id": "global-mmlu:b882d8461e843a6a3576b3442812f01f8f4c7417128b802b72d95864afef0e7e", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "f8ddfdc9a6d1d49bf880652f2ed862107bf63fffce076b64e605b799713523b2"} +{"id": "modality-text-global-ja-professional_law/test/789", "group_id": "global-mmlu:7bf905098b39d64fb64789728ff2cdb9dec8b198b7d87e5e7ded21fb33b355f2", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "07a0839480813844ba5abd01a74405da6d3669e30caded7e5238c1c12a469f33"} +{"id": "modality-text-global-en-professional_law/test/1020", "group_id": "global-mmlu:9d2e18865a02ac19dfb3642911a6913a0d76da04e325d9c5618d7d473132936f", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "334d7d3f74ca7675627564a73d9d2720acb3f25180f4315416446985cb211621"} +{"id": "modality-text-global-ja-miscellaneous/test/526", "group_id": "global-mmlu:bf91c9f673cef9a8b7306620b483a70f60d4e073e8518350c39c217dba0df136", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "c89265230bf442dc4f4e74d54104c6bf5fdc0830a4c250e319d7dda4fc98c83d"} +{"id": "modality-text-global-en-moral_scenarios/test/622", "group_id": "global-mmlu:b33c2fb6c1dcf29d5ca499f9d243b38492a41bd35f5d4d3dbdd0209391b57d5f", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "3c603908b8b0aa12ba755c82079b69048fa5002d250f329faf7502eb4269ebaa"} +{"id": "modality-text-global-en-virology/test/131", "group_id": "global-mmlu:59aab281f7fabd0e1e345a6dc4ce327d395481eb762f6ce258176db542abb211", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "86dc336f15cb3ba2b0cb4faf46de709a830dce18cc2770adf1bc2d6411d3c0ec"} +{"id": "modality-text-global-es-clinical_knowledge/test/193", "group_id": "global-mmlu:44d049417c9217aee54a984fdae2f2d8317c7ba8a1a906171043cc0a0c74c128", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "fe057f6f109e24f9c483104d39da89aba52dea1ffb526fbe8190f098f6ee147e"} +{"id": "modality-text-global-ja-clinical_knowledge/test/32", "group_id": "global-mmlu:db9b327aca575431aad095a3681d713c59639cf2b271b85599ed91e14104cad1", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "ee744b2a74573c15447aa3b3b993d8c13220c45a8408c0b10d0cede8edfc9d10"} +{"id": "modality-text-global-zh-moral_scenarios/test/62", "group_id": "global-mmlu:4a322d7d02492fb0fe6f5a98581dca53b7223f2ddeeb7d550783fe0797bece7f", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "4adabef52ccc4b976cf507ebbacbaabb5f53aa6331a47ce6dfea007b0688733f"} +{"id": "modality-text-global-es-high_school_psychology/test/265", "group_id": "global-mmlu:961454bc3f74765daf7d05b84cdfaf2deef0a39bf0e9773e56f55d5d99bbb650", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "2af93abaed65917b2b55ca9da0dce214bfa779d2cf0c19e0bffa5f94ff2c3be7"} +{"id": "modality-text-global-fr-electrical_engineering/test/90", "group_id": "global-mmlu:cc73804ba7b429ade40fb5cb5f937fabe27341f7cc5cfbbee4a1b9bea6320af3", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "e798f08f2fa012c8ac5d61b61e2711adda402a93211350fbe8702469a0a928e4"} +{"id": "modality-text-global-ja-moral_scenarios/test/180", "group_id": "global-mmlu:4a188079b332809a161acfbbb9493d1602d57adab5f0e330204c011bfe9791a1", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "031821ae0e7926f73ee73c9d487e4cc5ed71c5a66198c9cbf595f6580a722ebf"} +{"id": "modality-text-global-es-high_school_government_and_politics/test/100", "group_id": "global-mmlu:9302f48aaa3371bcf4e43be3f8016f2c5fdadb319ba972d426b70eb6dbc50578", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "70964a6206002311989f13379d4bfd2be6789d1b38d74a302617a4a377836e20"} +{"id": "modality-text-global-ja-high_school_microeconomics/test/110", "group_id": "global-mmlu:90c0ffaf13978d6849dd6e5f66b67ccbe5d17dfb9eb36557aad03ec1f57f58c3", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "c585a8ae6def992953f97c9ab676f651017c6f1fa3d7dd986b56c3c4c60af126"} +{"id": "modality-text-global-ja-professional_accounting/test/267", "group_id": "global-mmlu:bac2327c91468fb5a11d36d4bb1ba6fa4f9d60432ef3ffc58fb7d85aa3398d2f", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "79c9a22c9da24781795fee77ab4f0cc64cead23f4348ad62d34d7575fa4d6a77"} +{"id": "modality-text-global-zh-high_school_biology/test/259", "group_id": "global-mmlu:62fa88ab02935b4c2db77d188524373cb1d59c09caed73c90bd35064e4cd7b09", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "e9167ac681ef8f81940d87c203aaf9175e7ec009c5cafa7cfed05b467fbffc91"} +{"id": "modality-text-global-es-moral_scenarios/test/730", "group_id": "global-mmlu:a0478c135e0a361dedcb1140a816422f78c3d8b8322e60241426ef72e51371a4", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "9e5a6092c8e54d1b00046d36001345a00090da8a79aef0ef5a7a4bcbe00230bb"} +{"id": "modality-text-global-fr-public_relations/test/50", "group_id": "global-mmlu:f7c3b5ec3261ceb5fefee189137457e238cf6bd0c4bd5cd21c6f050db86e8261", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "38e5b224ecba051bc54569f2d607baebd03a70fc3c5cc01c3beff093f5668ed5"} +{"id": "modality-text-global-en-professional_medicine/test/69", "group_id": "global-mmlu:e140ff4635aab22c1c8af184bb3424235c57557a649faf2b8360da3d0681dbe9", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "f807403e9e4f7d816087cdfd7fa18373b471d2b3ec1055ed63d63a44a0f306c0"} +{"id": "modality-text-global-es-jurisprudence/test/53", "group_id": "global-mmlu:4fa53ea633ee78512fa644db8ae2dcd39e596dc6647fa5e48c8e59c7696384e1", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "3be5841558d342ff1e70334566abc643c4c368d7c48af59df531c4b61d509182"} +{"id": "modality-text-global-fr-professional_accounting/test/271", "group_id": "global-mmlu:3ee17a35e2cee4176be7e888f99e9c8f367067f278a7d0a0f30d9e2f73516aae", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "df67e004f3799f514e2d419909c1af4cc401552c97d47318b6adf19bcb3e8553"} +{"id": "modality-text-global-de-college_biology/test/137", "group_id": "global-mmlu:1da29d2a56afff2956354fda349563acf0c9a3b1fcb4c29794b6e5973f3c2eca", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "ca6365bf794fe745300635859c9fbcb31393717f6aa46f1891a8eb4566812574"} +{"id": "modality-text-global-fr-high_school_computer_science/test/19", "group_id": "global-mmlu:9306265c0827ee98c79e848d1e87ced7bd426f341e69d6298631706450191a9e", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "0c24bf00270b659981faffa649ea87ae32dfc510ddb6ded9faf60ea7b527fa2a"} +{"id": "modality-text-global-en-college_medicine/test/10", "group_id": "global-mmlu:ff0ff5717715efd3301a715aa528c3fd8c001f290e882160c6eb956289f882f3", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "0f43c408448bb8d0412f272f9c62767d30b45ee91da3e60bb91e91f3faf9379d"} +{"id": "modality-text-global-es-human_sexuality/test/70", "group_id": "global-mmlu:9649a221975025c8a8fc8223222d433bb3256cc0ebfa17b078f329f1c5bb6186", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "cef9909b22ee7c2390fbb4c6f8148939426d0ac7f93d65d4c7d6a3a6b96cbc81"} +{"id": "modality-text-global-ja-moral_disputes/test/196", "group_id": "global-mmlu:305d3743f17d43b863de1d43d4b68d96350db0dce53c7ce516be4bfab9583160", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "3707e27d6de5d72f095aafc3d29fbf534b52d429c069e8563244518a144c6f7f"} +{"id": "modality-text-global-fr-philosophy/test/168", "group_id": "global-mmlu:17f8f53f4b7f0e5421bf7e83e296d54dcb0ff79d99a092c846654e5bfcb38b12", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "c387bd107b22cf45f0a1d680a377e7c02636d679a687d88f9e095984c3f83bbe"} +{"id": "modality-text-global-en-high_school_psychology/test/265", "group_id": "global-mmlu:961454bc3f74765daf7d05b84cdfaf2deef0a39bf0e9773e56f55d5d99bbb650", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "6568e8c4dca1fef89b473466e83cb66184df8c7af4d634647aad73a2f0ae2556"} +{"id": "modality-text-global-de-moral_scenarios/test/767", "group_id": "global-mmlu:03b799e335f81510724761a482d9f5d6fa188bb3db9fa96ccf5a49209fe8c00a", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "60c0852e8cd512f4bf18b964ef9f71124107933220704f9e9d13f1955c3edc13"} +{"id": "modality-text-global-zh-medical_genetics/test/49", "group_id": "global-mmlu:5929ef368a92d7ee0f9168fe6e578f06fecb079b68367da512f87ee29f453936", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "ac88541230cfd5b330387d3ae620d63e6c6baead721a289b1a9e59909ec4ea71"} +{"id": "modality-text-global-es-high_school_geography/test/142", "group_id": "global-mmlu:ebde3a3054cbdb6c3d63622c004bfdff2415708f474ca91787dbd80def4cdc6a", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "670020467ac75eeafd9626629a4d2581874314a28560d6b3f9781ff93e8b4b58"} +{"id": "modality-text-global-fr-sociology/test/71", "group_id": "global-mmlu:a351cbf18b116234224592939bad0ee6c3f41112adda4e95c67ad7228b152e91", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "19e5cb3ca198bf017f47d8643c524bdec2fa068de7bea58fedab13cc94e15c9c"} +{"id": "modality-text-global-es-moral_scenarios/test/480", "group_id": "global-mmlu:313634799f463ac2fff7d56ad4be7ff37d326cc4c3ded54764afffd8b1466a75", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "83087356bb73c21ef4889f24092ed5fe2751bd730bcc41ee107b4fd88fa610e0"} +{"id": "modality-text-global-ja-high_school_government_and_politics/test/100", "group_id": "global-mmlu:9302f48aaa3371bcf4e43be3f8016f2c5fdadb319ba972d426b70eb6dbc50578", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "a1e09358894b2f453e75a97c75591571752fe4de2ec75b78c205c4f582e895f4"} +{"id": "modality-text-global-fr-jurisprudence/test/42", "group_id": "global-mmlu:aee8e65ea00728828c6407220286831af90630f3ee125a4db13a3a916a2cbbca", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "990be7da60688768ca556f33c9a7773c4cce6600c1f922bd33df72aa7005b782"} +{"id": "modality-text-global-ja-clinical_knowledge/test/84", "group_id": "global-mmlu:62449e8b84ecc8c5c8d2bfa5b49762db1b9de870a18b8868d5815759230af955", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "120cdd0ff8583a3555416f37a6021d5dd489af3d0607d18f549ec5acfda5afe4"} +{"id": "modality-text-global-de-high_school_mathematics/test/45", "group_id": "global-mmlu:a2ab370f4f6833caabdcbf6b962b734ea162ba7bdebd42b02f202261724e7493", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "835c091e68c6b2f77d120383ad48d1f001a78302fa64507ebcf9e0f1af7b89cf"} +{"id": "modality-text-global-en-professional_law/test/1383", "group_id": "global-mmlu:7eb9467697c98859e76e4618ed3b7b4b4dbe306aa4c6944928e44ed90ae1f934", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "d674361aebc5208ffadaf77ea43b1518ebde9490938abf38c9a3fae9c8095ac1"} +{"id": "modality-text-global-fr-philosophy/test/139", "group_id": "global-mmlu:299d406a95fba25c3eb18a7497cb01be478fc45791e454c1c8e8ff116ec5f090", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "27500bfbdb2b921fdac7ff212fe151a012fbb1aebe96461ebdd01c6262917b85"} +{"id": "modality-text-global-en-professional_law/test/1189", "group_id": "global-mmlu:2ad7e7d8e36ed6b270a4232fdd7d139f49ca92872378de46a439f7996a538e2f", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "e4ee52e4ab99be9e385d8f4c831e25a3bf0346c25ea89f4771dd2d8b6411b8c5"} +{"id": "modality-text-global-en-marketing/test/90", "group_id": "global-mmlu:db59c5dbe6e2ae3343097cf598b5b9e830b1ac6a11e0c8bd8fd2dc97b8729b01", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "dabd7f7a544371ece35accafde3b7d3912f9e5d666939c7e01c0bb9e6a0a8be6"} +{"id": "modality-text-global-zh-sociology/test/14", "group_id": "global-mmlu:eaf18143f999bb394db9b8601aec3ffa5c386d57e8a6deb5a901beb1ca054e3d", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "295d3cf09b3397a4380c5c32df9374e9e06a7c6be640aafb9a2993e1f9737573"} +{"id": "modality-text-global-zh-high_school_government_and_politics/test/58", "group_id": "global-mmlu:5c81f441c6f809ff256903eae4e3bbdae6f0286f36c274f3e099a16e718df831", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "99acf9ed23cb3b677d9b368447e061cc56f036cfbf83f9a3dc222eaa869dd84e"} +{"id": "modality-text-global-ja-nutrition/test/301", "group_id": "global-mmlu:7e9689610770055dfac707b537c5b44a6a4451f62074fd7e599dc2dd3e5f3d5c", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "56fbd932bb60ed66ebca0a0640cdd7a0b699ec78b0a66f45525030a067c5c1b4"} +{"id": "modality-text-global-fr-high_school_psychology/test/83", "group_id": "global-mmlu:6dd4809b7318857364f1937ec9522c2a52e9fc9143d129cf670e49b922da4c82", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "140f56657da6640e1f0960c63f41272342e94e8267c4be832303f3196eda8f38"} +{"id": "modality-text-global-de-moral_disputes/test/200", "group_id": "global-mmlu:71cbc88404f9175872c1878343646e63133d31f74637944f7b26945f0959ec15", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "52a3a212ea35b27cbb6be1ca483c4a498576e5e2cd6bbc46aadeb830f7426823"} +{"id": "modality-text-global-en-philosophy/test/104", "group_id": "global-mmlu:541ed884ce52d9f15a3777ecad176a77b28921df466dd26e9b938100e26df846", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "b8445c30779b98ae88bba350577a1d8ef4f7572778e0742c2802ff7dbe1f937b"} +{"id": "modality-text-global-en-professional_law/test/774", "group_id": "global-mmlu:c5b0ebd9049d79805a5ab17a316be3954708cc5e8ea8ac883605486c63a8e367", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "9c5acac1d073d9915d190343c9a6a1d2c4c542728808d427b0cb5fc289c54e76"} +{"id": "modality-text-global-zh-virology/test/85", "group_id": "global-mmlu:4cc9c2ed278a0d2fd8b4ec739ed91d322d9a07e84066364e54d42f6eea82d668", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "15e7e6e882200f9aeb9cdd78688a7fe52cc1914d014462f1b971db35f4511b35"} +{"id": "modality-text-global-en-professional_law/test/261", "group_id": "global-mmlu:1a2c91f04eaebf6fc47b7e0cb061f0752b9b42b36afca5da142cd960bbc24361", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "febb92153d6e3b8d81b22e0c7fbe5b39e1d5e3bd76822ad9cc290edfe4912c99"} +{"id": "modality-text-global-zh-high_school_psychology/test/456", "group_id": "global-mmlu:16afbe1a2a9c34783925ba31df5e39c85ebe9c4b458fba94a6b92e06acb7d788", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "79b97624aa0550c66de64cad2f0724916079d9483b4c94eaeda0e1cbb730c78d"} +{"id": "modality-text-global-es-moral_scenarios/test/153", "group_id": "global-mmlu:a1d0d823bb8e8bebff18f0cc0e427fb72def0b6c92bb6cd0a243d68a10178771", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "15dd0e759954759d4e7dd6eb2fb9a0489df26e9ef7aa7c9bf3b8af071ec42517"} +{"id": "modality-text-global-en-professional_psychology/test/128", "group_id": "global-mmlu:7394bd9bfcc64592a839f03238859933fdca486a661fc5bb4d490eba83cfcdc6", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "9fa568bd9403c270c2cbf4e2faf0c6bb4278590e9663230385250123dd2db1f0"} +{"id": "modality-text-global-fr-miscellaneous/test/58", "group_id": "global-mmlu:26ea6855132c0ef882cb69c9fa1539a009c830127c677a246e93ecf7301c2b79", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "b794cbc25409eefdb1320ac14fd2ccc5b9db5529aa5464ea8f73697d2a475cdb"} +{"id": "modality-text-global-es-moral_scenarios/test/786", "group_id": "global-mmlu:8400e5ab7ef8f0dd33deee2c9f016ab7a26651c69c28784c9db7c68be110f2f4", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "51f3b2aec208ca4d6ae0e4f0e232b037d2f5c919b8c98e71b5c8901046e16c68"} +{"id": "modality-text-global-es-moral_scenarios/test/249", "group_id": "global-mmlu:809e39758d444e7dbb3f748e32721744edc7584315e502f060db7b4005f223fb", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "ec62b3afa714de17bf8548bf2e1e5a182cabfbf8d108e5bd15ae1c3406e9afdb"} +{"id": "modality-text-global-ja-virology/test/107", "group_id": "global-mmlu:bef3d7084f33ecf5bf551857d702c72b8a292cc5726324f7a27030f11d91f05f", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "b4669c4a84f90918eb8acc0763c1118011149b61efca5a2f7d91744deeebbe79"} +{"id": "modality-text-global-de-high_school_biology/test/47", "group_id": "global-mmlu:3bfbcb867da8445241755451334e9bb186bb0a16642638fd05adbf3285aec58f", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "7f38f0de46b020e83b5afee77090ba7f3d4d3e08ddf840c4359b7e13a1475e77"} +{"id": "modality-text-global-es-human_sexuality/test/41", "group_id": "global-mmlu:b0451c93a9b92b9f4d297593e9a3010d4b335397131bb02f97c97b7dea6e16bf", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "30dac5f543f479acae4613393dbe9220dd2cff39b05d7159e24445304974641e"} +{"id": "modality-text-global-ja-miscellaneous/test/423", "group_id": "global-mmlu:804c79cee42c67d062549498a9672aa298492c058b49fb26ac70a0c27a44d626", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "540ef765a3462ebbd45845e57e52c5d361a0d6ba80435c43fd98165863fac742"} +{"id": "modality-text-global-es-high_school_government_and_politics/test/114", "group_id": "global-mmlu:2c773d052347a784ca2977e1f49941d98a05cc2fbce847862308a5a4014415ab", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "ba799ae950da7ec443d518007395e79fb0df970be34e4485022d5cdf6d80f9ff"} +{"id": "modality-text-global-en-international_law/test/33", "group_id": "global-mmlu:1eab75c34d1b692f4ed8ed5dd3999609b91dd413eaf1f146e712e48916cc59f1", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "04872abf19e2e5e6c61899c6593de2ed6ccb00f466d87657ce204125757755a1"} +{"id": "modality-text-global-zh-moral_scenarios/test/155", "group_id": "global-mmlu:a32cd20f714a895dc6b8f7a1f705ac3d5c0d1930d93b3a23ab9c14596e1e69d1", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "1a4be303847d218f67c37ea1f39b336ba78e060e1f3c8981d09031bd6718ef04"} +{"id": "modality-text-global-zh-moral_scenarios/test/274", "group_id": "global-mmlu:e5599eec163884fa120397719edd82058cde36ae66f8921f790676df6b74f094", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "070f8d88a40393937f16de48c690878ad0b99b12bea505318f21cb141a6375a5"} +{"id": "modality-text-global-de-professional_psychology/test/512", "group_id": "global-mmlu:509575fc59354879a3d48203e714ca4d02f739baefac0cee711a43ac52bf5e9d", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "29bd1b4b73dbfb3d89742f9fbb48b4cb3fdc1af1e6cb2cebabdbc7e205271768"} +{"id": "modality-text-global-fr-moral_scenarios/test/240", "group_id": "global-mmlu:8e30749a491e581ef570566203900b3a26735a1320cf526fde7d5ef2e8d1f017", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "858dc66767581fce54102be49ca7b4578661e14ca74d9d055633309e9764d559"} +{"id": "modality-text-global-zh-moral_scenarios/test/281", "group_id": "global-mmlu:a583638f3e8a223769e7f4d1c77ccb24f1af51ca5b454fefe1fc630a87645bcc", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "d4601a7fd7f46de42e0999547a6d599fa4d28a0d4ab619fcf2e0c976a46849ad"} +{"id": "modality-text-global-es-professional_psychology/test/80", "group_id": "global-mmlu:9192d8d789e35415b923f7cd52d8719e52f10204623b5614d3463cbb65531b45", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "d155d1c0e51c0c47140589bdf7f20dc31e5fbe730f0218951702fe175f139fd2"} +{"id": "modality-text-global-ja-marketing/test/58", "group_id": "global-mmlu:c87c4d16684a64212b7b24134d80e1b729e3860f78c9957dc60422b79d016876", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "dce5b00f907809602fb5dcff42d4b6446b54881e9a9ab381ef813f3f6825306b"} +{"id": "modality-text-global-fr-college_physics/test/77", "group_id": "global-mmlu:db955f06ecdcf8d40edb4c2282b737e54daa678dda8bfe81bd97413fef2315c9", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "0e688f8bb22bbfd79e2e8d62b7e68e91f884b82abaf791e74f7a399ded05a821"} +{"id": "modality-text-global-es-high_school_psychology/test/367", "group_id": "global-mmlu:2cf3376cab7c58f8f56cd9b784e4e269aea6900cad5a1740d56c50838a2be401", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "cc6e7bb399f254993f66a1b7d150053b5e54766cd4518955740ee0ae4d019a73"} +{"id": "modality-text-global-es-high_school_microeconomics/test/153", "group_id": "global-mmlu:cba22768aa8c57d542d5ae9a3803594c9f3b2ef18262046855ba2f766116cdef", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "411a473ff494de165aac2da2e787cdcc88c26bbbd98856a572d1c9b82c7398e7"} +{"id": "modality-text-global-en-elementary_mathematics/test/366", "group_id": "global-mmlu:5c3917d3d0d076201fd2d9125a573605522d367f5215663787dd4f3b32806869", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "3c6aaef71413fc8b49e340ce8080cd5266150e8633b4504a95e383699c33c4d3"} +{"id": "modality-text-global-es-virology/test/163", "group_id": "global-mmlu:2dc8ceae679195da1703b7c6069f0240313fffca07808db579f2dbc967d8be78", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "2c6bfd10a53a984b04a970e695206e677cddb043480e16b749a857302043af10"} +{"id": "modality-text-global-de-high_school_statistics/test/198", "group_id": "global-mmlu:5e3345de180c62c139135beede5b0a47c190ade0963562ae3b0422899341a807", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "47171b67f55cdac8f933aedb2f06b7fa4156b4209af05e5398a1de5225550e48"} +{"id": "modality-text-global-de-moral_disputes/test/36", "group_id": "global-mmlu:b76c4e8d1b3d6544c8e50a55f9f93fa7fa6e28284e10723f86e0d211bb762bfb", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "ea965daf27b466bf02636f4a8b9230dc74e39a89f698d9fdadbd791051ab879c"} +{"id": "modality-text-global-en-us_foreign_policy/test/77", "group_id": "global-mmlu:a9676017a07b9749458bf018b7501526d9b7513d0d58eb8a33223e83a4824e77", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "6f81ea1a59980f728e4a6bb732f1e6689575d3a4bea3cb8f39288e26e3872787"} +{"id": "modality-text-global-ja-philosophy/test/39", "group_id": "global-mmlu:9c296371a1b4efd39f70ec587b9dfe6cb0501899c958bd7f7dd45a0001d44343", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "c988c8cf46829f4159e1919974f0369cbab366000c225691a097d1aea8d2d40d"} +{"id": "modality-text-global-fr-philosophy/test/142", "group_id": "global-mmlu:84d1e9afeb520c81cdf181320658e5ecf6caf0a15a3a8a3ede0db0418f6ff972", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "41d01a0c502c9900f34ca517395c2e459e285deea29ce2f5ce0da8a7258c6ed7"} +{"id": "modality-text-global-ja-high_school_chemistry/test/33", "group_id": "global-mmlu:3cd6ff7ad4d9d0832d28200a363e5851779edbee52b66f489273c5ed3283ae3f", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "3efcf9f788a8b768b73e7c195620a6cd3532bd7c272e4596c7975e9493ada35c"} +{"id": "modality-text-global-de-high_school_biology/test/166", "group_id": "global-mmlu:80e4f44b943b329a63c648e8f407d4a15910816943cf3dcaa2f7c94724d2aa64", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "397a51b0dd27f34ffdfaa46d7342381760f691d7d1cac09f171b0b54c113e741"} +{"id": "modality-text-global-zh-public_relations/test/66", "group_id": "global-mmlu:7da184e5c4d4132cc8b673605cdbb137a1c4d48fa1802a2a762b2ee800d727d4", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "b671ac2772c34c7c9fa25fd9e2a1bd4a36214cda8bede4909dbf89b3276ea998"} +{"id": "modality-text-global-ja-moral_scenarios/test/295", "group_id": "global-mmlu:f8fee8abb51e944697d49cd3c529d2341fa77b7629d3c176856fe30e92f10300", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "967c8414fd1f35af996a9dd0ea575b6c351dca1d2e8c22289e5841c657f544f5"} +{"id": "modality-text-global-zh-professional_psychology/test/472", "group_id": "global-mmlu:eb4388976ea0ba6ce8123df10e6826b9f289ed15616aaa5ce132bc10f4e4294d", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "7318af8a12ffb03a615e6f7583a373e43d9e29fd6bfc0c0a49dc4845800321e9"} +{"id": "modality-text-global-ja-miscellaneous/test/14", "group_id": "global-mmlu:ae0fee8b38dd0299117c71c69dd045f4a23a9279168cd3e8259ccfdd1a89ee7e", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "66dabaa8c3209a9259eec3645a2bd47c8e6b367bfe4089c03281a56d70f8ac9f"} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/274", "group_id": "global-mmlu:a8f999faf2427bb9cea3fbb157a25c85e0b7fc5607c475b95166d2e23485257d", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "583266d4868c90e9d22bcdb630bef7641ff1bb859fb08c562f9356e1507f5052"} +{"id": "modality-text-global-de-miscellaneous/test/728", "group_id": "global-mmlu:ae38faf9551958ac76c409cdf38b07100522e27f8fc1e39620ca9719862a9245", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "c18380bcb52b9becb4f59f1ea5f3d9419c16e3564504a5cd3863f6fce8d4a9da"} +{"id": "modality-text-global-es-clinical_knowledge/test/80", "group_id": "global-mmlu:9e5f87662fb0e0c41eb6cf1c3ce223e1e8e28ad7473282565aa7c918b5530b0b", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "ce4eeb61cb09a26e407bc49876acb611af4a83d4b5e53268ad526ec3a44c236e"} +{"id": "modality-text-global-zh-high_school_macroeconomics/test/110", "group_id": "global-mmlu:434d8487d33f300cdb30626e62319a8db182da294366002e96d3a7dbe83f8ced", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "5c7214aae0e95051142eb9a17ab63aa2637a15b86273645de21602f13f653445"} +{"id": "modality-text-global-en-college_medicine/test/73", "group_id": "global-mmlu:dd98f45b344fd28a558c3b9be97b65a205a8443b20862e7481daa29e1f034d92", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "eba7f9ed93bc8841db8b15f352132dc95e2352709b94cbf47a1286b9311a7c06"} +{"id": "modality-text-global-zh-high_school_biology/test/96", "group_id": "global-mmlu:d2757f2e26f24a6cfa1055cc6fe50be0f144630941971db42c1d9ac52fc6b248", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "189a87666f7f7828732a968d3b2dca19ecf7b5f1366cffa91058aa228c0731ac"} +{"id": "modality-text-global-fr-sociology/test/188", "group_id": "global-mmlu:65e6ac288f24949fe66be7f9bf6b1773ce590c5725956c54b78afc0e4d29a2ce", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "7ebdd851606edbb09dbfad0bc7fbc35a506ada92ad9e40a35cf89a6dd0c76a12"} +{"id": "modality-text-global-de-moral_scenarios/test/587", "group_id": "global-mmlu:1213711e309371be8348bb528850da177283183e3ad7426fc170e97b3ebe8dde", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "c43b295477243db13c60d829ecdfc0cebeb59e5eca2fff9aefe6c306e4281d24"} +{"id": "modality-text-global-zh-high_school_microeconomics/test/176", "group_id": "global-mmlu:4d91de7721b679a658f24bc7b8cacf3a5a8b45d9931c49ae3b1889b33efeca3a", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "3817427cb4fa0ca47b7f9ed9159bcf3a37bc6b9c4b528c6d3ecc5e7f9128a447"} +{"id": "modality-text-global-en-high_school_macroeconomics/test/354", "group_id": "global-mmlu:8b759679350bad06ac20740c7cd700229bffe27fb9b2151cc9900a009cc156bd", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "cb92cdfa96e84711de73969ea00bd29ae9a3d8e822d8e2cdcb463c68866572d9"} +{"id": "modality-text-global-de-miscellaneous/test/365", "group_id": "global-mmlu:166bf3bf68fed1ae90cf1264f08e1bedd9374d658a8c6f89f25372cc59b5e25e", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "b7c9c52da2c5968d21645aa4072fee36201bd99c36df2796acd0edb33f6c3c8e"} +{"id": "modality-text-global-ja-philosophy/test/101", "group_id": "global-mmlu:89cd3e75d928c32cd3e0fbe1c5820988b0c89fa83c8bdde5b12eaeb8eba106bb", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "992739a13cef7d7c61d9324f5477ac7bb9681676d2be6b3a79d0397d0d4661aa"} +{"id": "modality-text-global-de-moral_scenarios/test/111", "group_id": "global-mmlu:62421021dfb39e04074aab2354547424d4de86c0948185fd9bad7ce2cb5505f9", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "fdeaabefd67e20d44bd0bc098ae345fa14c99fd01a44e1517c609b23e2febc96"} +{"id": "modality-text-global-ja-professional_medicine/test/127", "group_id": "global-mmlu:096a60765f752bff799f16c7e7a87e575529057493ff0d78f101b5afa44882b1", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "539ce6cc0768885f11d9dd01a1e7760f634a5b3eee17e9f37f8230ed28b2be88"} +{"id": "modality-text-global-zh-conceptual_physics/test/10", "group_id": "global-mmlu:e3f296c34fa0a27460e249cb4aacc15ec5dc1bd71161adc43eb90e8886a2b396", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "00bc30eefb85979fdf4ebd6a6aa2368906c40d9caa4ec3556d1da5da9f19226c"} +{"id": "modality-text-global-fr-moral_scenarios/test/148", "group_id": "global-mmlu:73b0da96bb25c3fd17d7fba57b619027029fbe6d035be3d5575507c6a4f18ac0", "label": "AR", "source": "academic-text-answer", "language": "fr", "length_bucket": "short", "text_sha256": "cd197c3c17a818b64efe11ca929b0dc1de5f3972c58eb71830d737871fbb9a4d"} +{"id": "modality-text-global-ja-high_school_chemistry/test/88", "group_id": "global-mmlu:8f55b019a039f29fe6a3e0f7d71909fa61c8fdabeda990f5a4bf4de1d841e1a1", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "915393d7802b15a446375e127576508e7a87f0b9c56c1d7cc4c5c867a2e970d1"} +{"id": "modality-text-global-en-high_school_government_and_politics/test/117", "group_id": "global-mmlu:f4a2b1cfab218b81536f8bda199069948094ae2d6fd6dab42dba4f5c39b78f78", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "abce64de969b7796499eed5ba38bb3c10f9d3cca7b530a25ecd4736c455ef5f2"} +{"id": "modality-text-global-zh-elementary_mathematics/test/325", "group_id": "global-mmlu:32815c97492480796728398ec6ddbf6e803937a02f9b151eac2f9857bff77afe", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "28a349188b477b51b670e36fc806e097c569e7e2d74e423a858c693cf4f28f1b"} +{"id": "modality-text-global-de-professional_psychology/test/279", "group_id": "global-mmlu:55b91513814721be46bf72567b85e46406427ea05744f85f8b1afe2ff03d9399", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "d018e47cf12341a762924f9c1eb0a12efc64664eedce8da2fd09773d139f8909"} +{"id": "modality-text-global-ja-computer_security/test/86", "group_id": "global-mmlu:fcc60d46c92e9e04e190740ebdd48043b10ead829939e1d0688ac4a0cb5e2457", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "077aeb3be3caaf51dddbebd343b69383baa30efa90459f735124a6c97c9acd3e"} +{"id": "modality-text-global-de-international_law/test/8", "group_id": "global-mmlu:e1b80688f2102aad9a2792abf28dc4d864b441431eb7934d99783d0bd9856081", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "bb172cf56e5655e5145e856afacbb0d801fdb33c0ad39a58f346ab6b907a1d76"} +{"id": "modality-text-global-zh-high_school_psychology/test/510", "group_id": "global-mmlu:f1c34c8b8e94a5e688bb0401558319ac9c3e0ec05ccbc46798e4081b4c17ea7f", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "c16226f270cb391da23942d41ef556d2151b43023f7b0c5c103eac4acbbf77bc"} +{"id": "modality-text-global-ja-moral_disputes/test/232", "group_id": "global-mmlu:179b915f6537b25484611a6477b057a07449f159b2fe62c16795363cc6b2840e", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "4befe265a0f8eecd6e1f434c8f5d1cfd7229a492907f2918500c5f608bac6a7a"} +{"id": "modality-text-global-ja-nutrition/test/297", "group_id": "global-mmlu:dd654f7f0720a69b9483d438cf2a27e816d0ebfa100680767bc5c6d3fe316f23", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "bc0d4a84d5856ae2b5c7473a374718fb0f07f10a4b244029eb8068726edcf1e1"} +{"id": "modality-text-global-zh-high_school_geography/test/110", "group_id": "global-mmlu:1a8cf2180eff71d8eb5be7e7cf5ab2a8e28b248af1585aef9058a9fd44ed9733", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "636db4f79654874b22072aa3e7fa96f07eee16106f46fb8cbc92c2c7475c5255"} +{"id": "modality-text-global-ja-nutrition/test/129", "group_id": "global-mmlu:8e1c5dc3668f3f93e14b09add02da06dab3b0607c2bbdcd0823a948c1a73eab0", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "8761f9544e404c262d0e02923aa5651b22082d68e4d0dbd9e37d35ecf12c2587"} +{"id": "modality-text-global-ja-moral_scenarios/test/456", "group_id": "global-mmlu:72247de96f6e827c279e20787e6cd3c8eaa597bdf122ecb7ba1ae22aa73adcda", "label": "AR", "source": "academic-text-answer", "language": "ja", "length_bucket": "short", "text_sha256": "df468b18965ed3b5d8afa0f1bb6875a564271ad0627360793ec451ee6e968a1b"} +{"id": "modality-text-global-zh-professional_law/test/500", "group_id": "global-mmlu:c050db6f6c23773e1f2ae9c847491521fddb80f77874e82c2463f620abb91883", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "6a619908a5dae0b6610508743898e7abb7bc043a5a469710bceb644c34ebb8f6"} +{"id": "modality-text-global-zh-human_aging/test/127", "group_id": "global-mmlu:fa15e3457e6d2376a57ae84f26daf8ff5da762d7ac8b54fce5e30261012d05b1", "label": "AR", "source": "academic-text-answer", "language": "zh", "length_bucket": "short", "text_sha256": "3de98abc2ad9b275b2368e15a71d44ea00ea75942c38ea807024c1ac15ff934e"} +{"id": "modality-text-global-en-high_school_government_and_politics/test/114", "group_id": "global-mmlu:2c773d052347a784ca2977e1f49941d98a05cc2fbce847862308a5a4014415ab", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "e6facb3bbdce9a2db23b88209f527bad5955cf4da6a37910d4674977d9e13461"} +{"id": "modality-text-global-es-high_school_psychology/test/273", "group_id": "global-mmlu:b79a212adf2ff7d84bb12ec9897e2fbbf6df95706a5b3199ce796da35abd9caf", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "1ded289ddff2ae92a9b73533ba560f49322179184ca9b474a651e899c6d2c9f3"} +{"id": "modality-text-global-de-high_school_chemistry/test/124", "group_id": "global-mmlu:288b95ff181aef53bad3f89952dd263219b9c3265cc0d11087a47000934e7d2d", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "707ccf2b0d8ba528a77c6508c14d1176836142d1b0cb8ba3e988c1f7f4e65c20"} +{"id": "modality-text-global-en-moral_disputes/test/273", "group_id": "global-mmlu:2eb0299f3c3b621114cae5eb9391f2756ca757982565a986ea20e765b43780a9", "label": "AR", "source": "academic-text-answer", "language": "en", "length_bucket": "short", "text_sha256": "fa089bfb490147b978d4098169a741f81616729c2fd1e87d9e35fda6e594e2cc"} +{"id": "modality-text-global-de-high_school_macroeconomics/test/284", "group_id": "global-mmlu:6b28368b860a575d3a65546590dfdcca1532064e19d6ebd90315b7cdf833d2cc", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "c4e68260b63854c475c652e9592e5775ba9aeaa751b924745c75cccda6712995"} +{"id": "modality-text-global-de-high_school_world_history/test/10", "group_id": "global-mmlu:d354c26a1c48d383339ea9b93c905824e2dba0f3bd6c9ed219b576359f6b47f5", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "2757ed762978759569cf4cb15ee69cd949b62741ae4a66f7f02090b9eef506de"} +{"id": "modality-text-global-es-conceptual_physics/test/164", "group_id": "global-mmlu:f5553ec1fba5cd5848c271fa2493d7ecbe84a9ec636bd31abe34ef6c7daf45e5", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "750e8cd986a9e35d1175c15ed67b51a654da87d3200455819f750a7b5ed91447"} +{"id": "modality-text-global-de-human_aging/test/53", "group_id": "global-mmlu:f8ed8d340e030dea152c5a13f970343c03ad1578419ab163cb0883f39ce9542e", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "3514ae78d24296dd7c2b9f430b5578182ee8886f1c9797733c64038fcf66192e"} +{"id": "modality-text-global-es-security_studies/test/140", "group_id": "global-mmlu:dabcd44f5753d1bc73edd0df4020113877c3069ec8d9144920016f08f661a6c3", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "5405303e976b0f7d4bc31afad141b88b12552a8fdb1b13070ce514a9a9dd497e"} +{"id": "modality-text-global-de-philosophy/test/202", "group_id": "global-mmlu:387dbf97bbb7fc52fbbfcd34fd559d98a760c835c5a1dd12e6231272c668d6f5", "label": "AR", "source": "academic-text-answer", "language": "de", "length_bucket": "short", "text_sha256": "51d5d6cb0ef9489213f68e459aa6da3874899563daefa993f9fcb434706d9f46"} +{"id": "modality-text-global-es-professional_law/test/807", "group_id": "global-mmlu:ccd785ce8dcd300095d1880748d8b5e4d5461c83c8dc202eac97335ad4eaaf7a", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "da25e27aeb9b0fc88cb45bdffce08e561db77fb0e2f3e3aa88e48a4fe72c6c7f"} +{"id": "modality-text-global-es-public_relations/test/61", "group_id": "global-mmlu:7108a38fa3cb42fc38ff7f712ef8578e86d1e3a8fb07cc4b42dd635a9a19f87e", "label": "AR", "source": "academic-text-answer", "language": "es", "length_bucket": "short", "text_sha256": "ef672382c30da812cbc27142d30673674df1b4e2af721ad38e1096efaa12bf37"} +{"id": "modality-aya-review-eng-26", "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", "label": "AR", "source": "aya-original-human", "language": "eng", "length_bucket": "short", "text_sha256": "03da286646085682a776eec98bec1f6276c01eeb16b7e79b63dca8b34ec3a67d"} +{"id": "modality-aya-review-spa-18", "group_id": "aya-author-2986e8162ec136cdcfdd5402e3765f8fa028d4f1726dfec5e501a6c650f5adbf", "label": "AR", "source": "aya-original-human", "language": "spa", "length_bucket": "short", "text_sha256": "0677b4aa5e7a59336e9bbc767d8d8f505f05484fd57e3dbeb24eb1c6fe38fc49"} +{"id": "modality-aya-review-spa-32", "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", "label": "AR", "source": "aya-original-human", "language": "spa", "length_bucket": "short", "text_sha256": "0c3ee6761f0f49573fd1d2514229862a9a995ab6a0b4b97a7ae8a1a9659e855b"} +{"id": "modality-aya-review-fra-0", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "00085a70e04ad5f9619e6adc9212172e9a79f960898a897b7081c817fb41dc65"} +{"id": "modality-aya-review-fra-2", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "003400e3ddc0deae7888904482ad57f739923e4e3143c6fcb6047d940bdeaf6e"} +{"id": "modality-aya-review-fra-5", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "00f9adec6b3ac9eebd05015037fd17b521104c9275b4870f3b3564ec5f642933"} +{"id": "modality-aya-review-fra-6", "group_id": "aya-author-1245db41ea2e3a92042cbeb15a6dd7230ab562db06b9d6286e7342fd442e5fd1", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "0148759121374fb00b49a8ee304f3f27bdf79435f9cc45406cd2608997908f12"} +{"id": "modality-aya-review-fra-9", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "0312504d169c0bf9b05ee14b20153730dd96745c2a756d93cb4792c3f97e35a9"} +{"id": "modality-aya-review-fra-12", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "0434346f2f814767fcfe7a876813a9fb5049c68f1a6f55eca2463ecb78f15e5d"} +{"id": "modality-aya-review-fra-16", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "05205a83cdb9b8e5cc5e621929353de70417deff8a5f73c500e2b5d206854a3d"} +{"id": "modality-aya-review-fra-17", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "05b1e7c13a66897d5aef6c595996ea4f1467ca744051394d19528c3acee7b918"} +{"id": "modality-aya-review-fra-21", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "068bd805ec24cc6566029dc28b04e768625bb35d06fbb691aceb8c1bada4790a"} +{"id": "modality-aya-review-fra-23", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "06eaf7d8ae23ab4457660d109c9ad6df77e957f7fd2f34b1cb9a6367ade916da"} +{"id": "modality-aya-review-fra-24", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "070fb3d1b82428fa7cda9022f7454992ab5b767d85f30b7dee6863835cdecab3"} +{"id": "modality-aya-review-fra-25", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "0713779fa0514fcd6ab452fd8756c64685b7079450d0b7093fb130f36deaa87b"} +{"id": "modality-aya-review-fra-26", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "07625ef864193d8a8765c87fae1cbadc8ce7531cd64c20567a2829e76e8328c7"} +{"id": "modality-aya-review-fra-31", "group_id": "aya-author-1245db41ea2e3a92042cbeb15a6dd7230ab562db06b9d6286e7342fd442e5fd1", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "08a92893ee6537ed8f55161328cae4af44f9916f3642bf8782e07673684d8f34"} +{"id": "modality-aya-review-fra-33", "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", "label": "AR", "source": "aya-original-human", "language": "fra", "length_bucket": "short", "text_sha256": "0917568584175cd9da38fcc9755b1d1d1014d41bfe956793cfb41dc40b2972b6"} +{"id": "modality-aya-review-deu-19", "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", "label": "AR", "source": "aya-original-human", "language": "deu", "length_bucket": "short", "text_sha256": "2f6fd43cee6490c64023b0dc95f25b7171d754065bb88b30a70fafa34b38c230"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:head", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "e397d6ecbea9c3d12d5d8f57e0609c35c9fa2b89d242198f88039e2c2d26b9bd", "position": "head", "actual_tokens": 4096, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:middle", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "44ccc49a8f703445ba95bbbcab01c98d7d204e993b362c258800178d766782c9", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:4096:tail", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "8479f32abf88d59ea8c4b2ba4c5e4794e89f0bf574245ea304dfa3105e84d2f4", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:head", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "d0531915ffca17a9861eb1a9e0e8e31320fcd2fd2385f19fc2db57294b60424a", "position": "head", "actual_tokens": 8192, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:middle", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "8da8f50a402410ffcdbe7b6a47fdbf5138162685cc150d61cf7020648d49ac87", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:8192:tail", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "5108a19960bfc2035ce66fee4b1363f20e8920d279cc74fdfaf0850566afc1e8", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:head", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "fb15d03f5ea32aa271fa8738fef92634aa650616a44d9c9e512eb0ce00544974", "position": "head", "actual_tokens": 16384, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:middle", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "3ecd113541cea2484df95dc88d8164bab2f4d0a0421e00187543fe1d72cb8d64", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:16384:tail", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "b753b6f334e43362d6a993074ba29b2069cade3ef79f29e54f0c0caf8a3cb397", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:head", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "43a8200c4def64fafd4462a83e973676115e0591976a9f85f31956d04f617171", "position": "head", "actual_tokens": 32768, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:middle", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "4b8032988482f39d706046d406184c45e4fa49e1f2c5340548ab22c6af79816c", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR:32768:tail", "group_id": "diffusiondb-author-530543a7cbea53f43a2ca32e671c362fb9bbf636ca75ac4bc66bd9e7991d89a9", "label": "AR", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "bfeea80e91b4e9facc1b3eaddc56257f61c7385d522fedfccc906c21254877b0", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-00072dc1c2aec7b33a42913013ec3ac1fb8866d50a13335295054289992706a4-en-AR"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:head", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "125959c49e8ff2f2283aa3a963227e61e11677817e41f2cf1e7918cff179192b", "position": "head", "actual_tokens": 4096, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:middle", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "75d0bf61a9e588961a2751ef2efaf857e193f55ee23b91df12a83bd64ec6b265", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:4096:tail", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "eca791e09cdce5d2264d97e574d769697318da0e4f800b24feb6ffe6464e1c1e", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:head", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "8deae601097848ffc28818822997e7552dd3d243a45b888a3cc3e10304e540b9", "position": "head", "actual_tokens": 8192, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:middle", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "740f9aaae0059c80b681d1aed283ea28fe9ee9be99ee4c6bc29ac7ed76cda9de", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:8192:tail", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "bc28205fb5999fcad52bffa20971e54c0c44be80abe2b3f069999eefe86a142e", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:head", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "9d127a59f179669a3ea542b0d3bb20010a65e5915650267f4d779f9c89010751", "position": "head", "actual_tokens": 16384, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:middle", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "9c79eb12c63c44ea52666852a7ba0eaeec2d37a4b94149ea8bd1b9dc69425fcf", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:16384:tail", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "aa365f4cfccb0996b4806b57c2e0d84738ea69bc70ce7efddd156a0a2942fbe7", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:head", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "1d554862f5e2239a5b87353ceca17dd8b430a389b7bf2e52fd32c834ca94a2d3", "position": "head", "actual_tokens": 32768, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:middle", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "6acbbebd472a21b59385036b4d3a87e0af21923f4ac657a5c2faf8d6b05b4035", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH:32768:tail", "group_id": "diffusiondb-author-cdd0ff88d270b227eb4cd830aa43cb1af5644ee6b3f1737e00583b69078c0af7", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "fd4407d9abf95e3f23bf869afaefa7ba6f79f9bc32976eb02f773f5b9c815e36", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-00085271bde738d0fa52ac6b1e2fd112f8856eb0ef591b8cc130e3929dff46d1-en-BOTH"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:head", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "a42b3c639c476b0c1c69a88b08d31eaa5fbe4194bd6870bbd224dc9693467fb3", "position": "head", "actual_tokens": 4096, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:middle", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "d7fc9667af55b087f47b09b416d126becc03540a1cb4bab735c70ec141d5ab19", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:4096:tail", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "4096", "text_sha256": "48c222a2a0929d1a32fecd095888f252fdfc151dac47368189ad5277be42f70d", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:head", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "82649eafda87eaa4218c032f9719afe68644309124526fcf9ce0d5c98cdcf303", "position": "head", "actual_tokens": 8192, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:middle", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "d5878ba96eb3245dc5b250fa354fb8bd5c075676d99fa1a78dae9ebdd43584ee", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:8192:tail", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "8192", "text_sha256": "db4c6c40971d2c23209a4fe060f04b904dd04b5581befc507838b5cd9a5f008b", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:head", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "3170b77c5da3d647f8cedb18c38bc7fde416f319821906be30e898cac605f048", "position": "head", "actual_tokens": 16384, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:middle", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "401756fd8ef77bfb7b3d53adeabacc59f9d0ba9adc75f510d166fd79ea555750", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:16384:tail", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "16384", "text_sha256": "afe7a06a967de8fe5e79e3b8aa94fd36ae6d3a0b3c3b6bad83570699a04107c0", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:head", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "41c2e98e9bdc96ae9de8639686ab0825201d03b1038429f81d4af177b8d64609", "position": "head", "actual_tokens": 32768, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:middle", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "dde87433055a3be760e107aabcec289ab9f6c0a99a26fb81841b5e08f0945a0b", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION:32768:tail", "group_id": "diffusiondb-author-1dc164876205374e5bea8cab0dc952b8ffda8dcce9a9b09ff88f3bc62daf4b4e", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "en", "length_bucket": "32768", "text_sha256": "0424b79d9aa2d764f6ee80c63c1f80a2a1f4a4d795884bbd6ebc7d1b3686a73f", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-000fcd9c092f4ec840c464d8c8464467b02763a17f997ad279e799723a6f9a17-en-DIFFUSION"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:head", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "acb8962985fbd8982a67661ca6f0dcd9201fe9c5f1e7ab7a59e614000f928d87", "position": "head", "actual_tokens": 4096, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:middle", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "da447d123d1709349540ceb7192c896cbbb7f4060aaad50c66464299ffef2b61", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:4096:tail", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "11ef5c5f3c69dd0d8b63cf2ff47c150fab01ede911638795b8341af11fc1d899", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:head", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "95a821dd86b1d6f26f3a9e7aa570678d1a2b84fc35b6273474a35253a7a1d36c", "position": "head", "actual_tokens": 8192, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:middle", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "7facda575a47cb7791dafebe81132243fc04d61fb13b84363af2fc83e9f9a1cf", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:8192:tail", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "87460196c2b2f24660611abeacf47cfe826e5fbb4d9beebe8011f0fbe88d7c0d", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:head", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "f0f8c373202b826e6b902f7df5de902743e813ea03e98562a7d3242eb6b6879a", "position": "head", "actual_tokens": 16384, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:middle", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "8b2e36ef2a5cec7f8be9aebaf239c521ad8e4ecf041926f26740d836f5d48378", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:16384:tail", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "c96d98ad0f7060ecc24a11f89c800a4653d6230184117414898dc5aec07a61fb", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:head", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "f54af1f215154ce2727d07f4862cba14c069a691e92e22df2386aae1f6034345", "position": "head", "actual_tokens": 32768, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:middle", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "869f0212020c470703e959cde5cb9f65dacd54d8aa41de28c0d5815912b2183d", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR:32768:tail", "group_id": "diffusiondb-author-775b1b4bd1b0c265d37ea8369580969804f90590b1d79fba073290b22cafd6c3", "label": "AR", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "4a14a9bc5d20137503e868a82415510123dbf8d3287d6a44ecefe021a2389e55", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-00121f6069a3ef9b8603ff0c434d5b9ff2e5b7fb40129b1c5c9e7ab5755abaeb-zh-AR"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:head", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "b68720b113a7e21939a79a6d67e6469be6c95ab2655558ae3cbf33f0713e0139", "position": "head", "actual_tokens": 4096, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:middle", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "f27d7ad357ddd4c65089ec0d3d75456d1f47642a06f9fca406e1f8be3a0a7015", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:4096:tail", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "9a1d521a93f5bc1a06a5230122358d7546e752a507b43081bfa64dc444b9b99c", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:head", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "9fea0e37a48ea73d9d0e772920392b5a32f3bd209d003f592b479d2ec03b75d3", "position": "head", "actual_tokens": 8192, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:middle", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "217b5e5bdc6546a0827c65c4b4aa5b34f9bb0206c60c8c8d19d667646bc94122", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:8192:tail", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "1230d48f3c85a8cb749773656ae76339c4dbcc3bec4aba0e8c3964020cf97bde", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:head", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "dc83d845fccb20e7e6b45155bd6d058b4fd9f42c104ceea643d2a184985adec9", "position": "head", "actual_tokens": 16384, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:middle", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "cfd77cc81cd676951d5cd70e918ce047e23a5fcdcd2fce0fb04f55214340e868", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:16384:tail", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "a660dcdb0175fac2a14e003979c8c01a6182d35167c6af165cfbfcc4193907e3", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:head", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "32f9f065e77f8bc9d31d5bfe945d0d10b73918d9c045b18862e9f6345b062452", "position": "head", "actual_tokens": 32768, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:middle", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "7c4982126034a114745702a7ac9f34e1162196bc3aadc2e3da4540e9949e5ac2", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH:32768:tail", "group_id": "diffusiondb-author-f43294fcf1eb88f650ca1fd4c2ee95a73b16bf989c95d8a9a0aab38d112f2309", "label": "BOTH", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "838671b52b093c5888ad6393057274a9d7d5e8d550b606acf1e60d9cb1745236", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-000088ee732e834031da127564508a978472996bb532a6c7ab8693e3e4fe0977-zh-BOTH"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:head", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "dcb020811aeb458dc9bee6bfbd51bc3dfd9100d2663ac90892bff9b7e647dd9d", "position": "head", "actual_tokens": 4096, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:middle", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "df6db723ab68ee5c982473ccfd9c776c44f7a65937c127408f0d959b85a6657f", "position": "middle", "actual_tokens": 4096, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:4096:tail", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "4096", "text_sha256": "689ea61f3400fa4b39c0576ca6caa8e7f58590bc1d63601e8536b6e16ba30205", "position": "tail", "actual_tokens": 4096, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:head", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "8fb6adc5c6bc5125a38d09caad8a96081c4fcaa2350d22779eccaf37f93c3c21", "position": "head", "actual_tokens": 8192, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:middle", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "279ebbe51b4a7d1d7fb81cba29d00d5f3aaa4da5a56f28d51bea8efd1bf54b69", "position": "middle", "actual_tokens": 8192, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:8192:tail", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "8192", "text_sha256": "daaafedd99f7cf984a6f1ae31073bf3a0a944d14c14517e1739b4c55153c8daa", "position": "tail", "actual_tokens": 8192, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:head", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "00dc636e1de622d8f715d11149246d5baf1413a2aa1cfe85bd86551e227caf1e", "position": "head", "actual_tokens": 16384, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:middle", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "1d5feca1ae2c93af42160e3a1272516ae7729d93b354fcad1a03e2bf2baabff0", "position": "middle", "actual_tokens": 16384, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:16384:tail", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "16384", "text_sha256": "fe7e4ee84d077d5a149e4f8946608e0d771ad34489d1d42bdd3a1fbf124e88b2", "position": "tail", "actual_tokens": 16384, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:head", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "db23affc665024e03a6efaeba2dbc4bfaccdd991d92d671844afbaab3c51d4d6", "position": "head", "actual_tokens": 32768, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:middle", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "5ff0aad6acb468ab6039122cfce6b75622ee2f57791e76ecb3671b792a5fec19", "position": "middle", "actual_tokens": 32768, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} +{"id": "context:modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION:32768:tail", "group_id": "diffusiondb-author-05355390e248f6678949372c1f6be7a094acad99c27518a91d6c0f05790ca3fe", "label": "DIFFUSION", "source": "authored-context-on-frozen-source", "language": "zh", "length_bucket": "32768", "text_sha256": "f99867f01da4bc2c824fdf1a0d9ab59c12f53ef61d28a4342c080a875efb2c9b", "position": "tail", "actual_tokens": 32768, "parent_id": "modality-0001439dedd74a51fb5a5722562186b556a09eabfd8802a56f20ac971da95336-zh-DIFFUSION"} diff --git a/evaluation/freeze-decision.json b/evaluation/freeze-decision.json new file mode 100644 index 0000000000000000000000000000000000000000..4f500d65dd4267fd4595512aae3522376e6b0ab4 --- /dev/null +++ b/evaluation/freeze-decision.json @@ -0,0 +1,33 @@ +{ + "decision": "accept Vela Base step200 after all predeclared short and long development retention gates pass", + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "selected_step": 200, + "classifier_pooling": "mean", + "vela_adaptation_steps": 200, + "vela_selected_step": 200, + "vela_retention_checks": { + "short": true, + "long": true, + "4096": true, + "8192": true, + "16384": true, + "32768": true, + "natural_aya": true, + "academic-text-answer": true, + "authored-output-contract-with-gallery-caption": true + }, + "retention_gate_not_relaxed": true, + "test_used": false, + "test_predictions_before_freeze": false, + "adapter_sha256": "52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00", + "final_data_sha256": { + "corpora/modality-v2/test.jsonl": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "corpora/modality-v2-context-v2/test.jsonl": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c" + }, + "final_rows": { + "corpora/modality-v2/test.jsonl": 1339, + "corpora/modality-v2-context-v2/test.jsonl": 72 + }, + "original_baseline_pooling": "preserve its published mean configuration; no pooling override" +} diff --git a/evaluation/migration-adaptation-plan.json b/evaluation/migration-adaptation-plan.json new file mode 100644 index 0000000000000000000000000000000000000000..eb84fe81bb90c9060e3101935380de8768481cea --- /dev/null +++ b/evaluation/migration-adaptation-plan.json @@ -0,0 +1,11 @@ +{ + "purpose": "bounded Vela Base adaptation before final freeze", + "steps": 200, + "learning_rate": 2e-06, + "same_frozen_partitions": true, + "test_used": false, + "selection": "source-macro-f1", + "migration_gate": "unchanged base-migration/modality-v5/plan.json", + "task_head_preserved": true, + "sampling_semantics": "same global4 samples, explicit token-budget microbatch regrouping; dropout trajectory can differ" +} diff --git a/evaluation/migration-decision.json b/evaluation/migration-decision.json new file mode 100644 index 0000000000000000000000000000000000000000..cb570e6b1ed33d1e473eabd13bdc13e8477f52ab --- /dev/null +++ b/evaluation/migration-decision.json @@ -0,0 +1,15 @@ +{ + "checks": { + "short": false, + "long": true, + "4096": true, + "8192": true, + "16384": true, + "32768": true, + "natural_aya": true, + "academic-text-answer": true, + "authored-output-contract-with-gallery-caption": false + }, + "accepted": false, + "test_used": false +} diff --git a/evaluation/migration-dev-old.json b/evaluation/migration-dev-old.json new file mode 100644 index 0000000000000000000000000000000000000000..09e7a88d3c79328efa86c1431e69d1b3ec68f7f4 --- /dev/null +++ b/evaluation/migration-dev-old.json @@ -0,0 +1,1410 @@ +{ + "rows": 1001, + "accuracy": 0.8551448551448552, + "macro_f1": 0.8591739356008521, + "macro_f1_present_labels": 0.8591739356008521, + "per_label": { + "AR": { + "support": 473, + "tp": 328, + "fp": 0, + "fn": 145, + "precision": 1.0, + "recall": 0.693446088794926, + "f1": 0.818976279650437 + }, + "DIFFUSION": { + "support": 264, + "tp": 264, + "fp": 74, + "fn": 0, + "precision": 0.7810650887573964, + "recall": 1.0, + "f1": 0.8770764119601329 + }, + "BOTH": { + "support": 264, + "tp": 264, + "fp": 71, + "fn": 0, + "precision": 0.7880597014925373, + "recall": 1.0, + "f1": 0.8814691151919867 + } + }, + "confusion": [ + [ + 328, + 74, + 71 + ], + [ + 0, + 264, + 0 + ], + [ + 0, + 0, + 264 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 180, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 180, + "tp": 180, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 180, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 24, + "tp": 24, + "fp": 12, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 0, + 24, + 0 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 720, + "accuracy": 0.8152777777777778, + "macro_f1": 0.791155643058494, + "macro_f1_present_labels": 0.791155643058494, + "per_label": { + "AR": { + "support": 240, + "tp": 107, + "fp": 0, + "fn": 133, + "precision": 1.0, + "recall": 0.44583333333333336, + "f1": 0.6167146974063401 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 62, + "fn": 0, + "precision": 0.7947019867549668, + "recall": 1.0, + "f1": 0.8856088560885609 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 71, + "fn": 0, + "precision": 0.7717041800643086, + "recall": 1.0, + "f1": 0.8711433756805808 + } + }, + "confusion": [ + [ + 107, + 62, + 71 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + }, + "aya-original-human": { + "rows": 29, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 29, + "tp": 29, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 29, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 150, + "accuracy": 1.0, + "macro_f1": 1.0, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 70, + "tp": 70, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 70, + 0, + 0 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "deu": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 186, + "accuracy": 0.7849462365591398, + "macro_f1": 0.7998805256869773, + "macro_f1_present_labels": 0.7998805256869773, + "per_label": { + "AR": { + "support": 82, + "tp": 42, + "fp": 0, + "fn": 40, + "precision": 1.0, + "recall": 0.5121951219512195, + "f1": 0.6774193548387096 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 40, + "fn": 0, + "precision": 0.5652173913043478, + "recall": 1.0, + "f1": 0.7222222222222222 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 42, + 40, + 0 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "eng": { + "rows": 5, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 5, + "tp": 5, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 5, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 150, + "accuracy": 0.7333333333333333, + "macro_f1": 0.7411255411255411, + "macro_f1_present_labels": 0.7411255411255411, + "per_label": { + "AR": { + "support": 70, + "tp": 30, + "fp": 0, + "fn": 40, + "precision": 1.0, + "recall": 0.42857142857142855, + "f1": 0.6 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 8, + "fn": 0, + "precision": 0.8333333333333334, + "recall": 1.0, + "f1": 0.9090909090909091 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 32, + "fn": 0, + "precision": 0.5555555555555556, + "recall": 1.0, + "f1": 0.7142857142857143 + } + }, + "confusion": [ + [ + 30, + 8, + 32 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "fr": { + "rows": 150, + "accuracy": 0.8, + "macro_f1": 0.8181818181818182, + "macro_f1_present_labels": 0.8181818181818182, + "per_label": { + "AR": { + "support": 70, + "tp": 40, + "fp": 0, + "fn": 30, + "precision": 1.0, + "recall": 0.5714285714285714, + "f1": 0.7272727272727273 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 30, + "fn": 0, + "precision": 0.5714285714285714, + "recall": 1.0, + "f1": 0.7272727272727273 + } + }, + "confusion": [ + [ + 40, + 0, + 30 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "ja": { + "rows": 150, + "accuracy": 0.94, + "macro_f1": 0.943391371472682, + "macro_f1_present_labels": 0.943391371472682, + "per_label": { + "AR": { + "support": 70, + "tp": 61, + "fp": 0, + "fn": 9, + "precision": 1.0, + "recall": 0.8714285714285714, + "f1": 0.9312977099236641 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 9, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + } + }, + "confusion": [ + [ + 61, + 0, + 9 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "spa": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 186, + "accuracy": 0.8602150537634409, + "macro_f1": 0.8705314009661835, + "macro_f1_present_labels": 0.8705314009661835, + "per_label": { + "AR": { + "support": 82, + "tp": 56, + "fp": 0, + "fn": 26, + "precision": 1.0, + "recall": 0.6829268292682927, + "f1": 0.8115942028985508 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 26, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 56, + 26, + 0 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "zho": { + "rows": 3, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 3, + "tp": 3, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "short": { + "rows": 929, + "accuracy": 0.8568353067814855, + "macro_f1": 0.8609653408729382, + "macro_f1_present_labels": 0.8609653408729382, + "per_label": { + "AR": { + "support": 449, + "tp": 316, + "fp": 0, + "fn": 133, + "precision": 1.0, + "recall": 0.7037861915367484, + "f1": 0.8261437908496732 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 62, + "fn": 0, + "precision": 0.7947019867549668, + "recall": 1.0, + "f1": 0.8856088560885609 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 71, + "fn": 0, + "precision": 0.7717041800643086, + "recall": 1.0, + "f1": 0.8711433756805808 + } + }, + "confusion": [ + [ + 316, + 62, + 71 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "unspecified": { + "rows": 929, + "accuracy": 0.8568353067814855, + "macro_f1": 0.8609653408729382, + "macro_f1_present_labels": 0.8609653408729382, + "per_label": { + "AR": { + "support": 449, + "tp": 316, + "fp": 0, + "fn": 133, + "precision": 1.0, + "recall": 0.7037861915367484, + "f1": 0.8261437908496732 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 62, + "fn": 0, + "precision": 0.7947019867549668, + "recall": 1.0, + "f1": 0.8856088560885609 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 71, + "fn": 0, + "precision": 0.7717041800643086, + "recall": 1.0, + "f1": 0.8711433756805808 + } + }, + "confusion": [ + [ + 316, + 62, + 71 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + } + }, + "actual_tokens": { + "min": 9, + "max": 32768 + }, + "forward_batches_ms": [ + 7039.483752916567, + 21.584214991889894, + 20.40059503633529, + 19.5585381006822, + 19.531563972122967, + 19.734615925699472, + 19.469924038276076, + 19.26641701720655, + 18.840476986952126, + 19.38274095300585, + 19.678111071698368, + 19.068480003625154, + 25.945398956537247, + 19.180268049240112, + 19.252500031143427, + 19.425202044658363, + 19.266484072431922, + 19.290908006951213, + 19.509149016812444, + 19.28546593990177, + 20.57167107705027, + 19.616086035966873, + 20.440079038962722, + 29.335040017031133, + 24.305583094246686, + 58.74593602493405, + 19.150480045937, + 19.049357972107828, + 19.0588979749009, + 19.150931970216334, + 19.140038988552988, + 19.038981990888715, + 19.242074922658503, + 19.93995893280953, + 19.138233037665486, + 19.876311998814344, + 18.998789018951356, + 19.22773791011423, + 19.245053990744054, + 19.192507956176996, + 19.094596966169775, + 19.125562976114452, + 19.157727947458625, + 19.24101705662906, + 18.754982040263712, + 20.657840999774635, + 19.407549989409745, + 19.27121204789728, + 19.349119975231588, + 20.35636897198856, + 19.340497092343867, + 19.41229496151209, + 19.209291087463498, + 19.234362989664078, + 19.478514092043042, + 19.862476037815213, + 19.685033010318875, + 25.732986046932638, + 21.024837042205036, + 62.87668400909752, + 62.32619902584702, + 61.78742798510939, + 62.28843401186168, + 62.4942280119285, + 61.87841505743563, + 61.24127807561308, + 62.97354004345834, + 63.574064057320356, + 63.25711507815868, + 61.819158028811216, + 60.69516693241894, + 61.04927998967469, + 66.55711494386196, + 62.53866502083838, + 64.0173009596765, + 67.36680003814399, + 64.94186201598495, + 225.73108307551593, + 221.4175140252337, + 216.55263099819422, + 214.4655380398035, + 211.94534294772893, + 209.52967007178813, + 207.52157899551094, + 226.30385495722294, + 204.03683395124972, + 212.17665902804583, + 218.6698050936684, + 214.6064389962703, + 208.55662401299924, + 205.77978307846934, + 214.26206594333053, + 208.822840009816, + 210.335728013888, + 212.93231600429863, + 848.0608450481668, + 825.0530819641426, + 812.0404820656404, + 817.2862839419395, + 802.6156988926232, + 838.2717769127339, + 831.2934199348092, + 788.4137380169705, + 853.9408210199326, + 869.962121010758, + 827.322592958808, + 890.770270023495, + 888.5154149029404, + 808.456527069211, + 845.0542960781604, + 794.9883950641379, + 803.5358130000532, + 815.9813659731299, + 3343.9996540546417, + 3326.254892977886, + 3468.5617400100455, + 3375.5812710151076, + 3486.4518040558323, + 3298.9523640135303, + 3513.150369981304, + 3348.251656978391, + 3501.9268669420853, + 3346.9400759786367, + 3471.5288559673354, + 3441.1905839806423, + 3405.6642989162356, + 3278.5228959983215, + 3227.0270970184356, + 3282.31126407627, + 3352.355351904407, + 3549.3063640315086 + ], + "precision": { + "parameters": "float32", + "forward_autocast": "bfloat16", + "device_type": "cuda" + }, + "data_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + } + ] +} diff --git a/evaluation/migration-dev-vela.json b/evaluation/migration-dev-vela.json new file mode 100644 index 0000000000000000000000000000000000000000..d23285d728e929508c16df4e79590b99713bec59 --- /dev/null +++ b/evaluation/migration-dev-vela.json @@ -0,0 +1,1410 @@ +{ + "rows": 1001, + "accuracy": 0.8411588411588412, + "macro_f1": 0.8456991326446053, + "macro_f1_present_labels": 0.8456991326446053, + "per_label": { + "AR": { + "support": 473, + "tp": 314, + "fp": 0, + "fn": 159, + "precision": 1.0, + "recall": 0.6638477801268499, + "f1": 0.7979669631512071 + }, + "DIFFUSION": { + "support": 264, + "tp": 264, + "fp": 66, + "fn": 0, + "precision": 0.8, + "recall": 1.0, + "f1": 0.8888888888888888 + }, + "BOTH": { + "support": 264, + "tp": 264, + "fp": 93, + "fn": 0, + "precision": 0.7394957983193278, + "recall": 1.0, + "f1": 0.8502415458937198 + } + }, + "confusion": [ + [ + 314, + 66, + 93 + ], + [ + 0, + 264, + 0 + ], + [ + 0, + 0, + 264 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 180, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 180, + "tp": 180, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 180, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 24, + "tp": 24, + "fp": 12, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 0, + 24, + 0 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 720, + "accuracy": 0.7958333333333333, + "macro_f1": 0.7650437660438248, + "macro_f1_present_labels": 0.7650437660438248, + "per_label": { + "AR": { + "support": 240, + "tp": 93, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.3875, + "f1": 0.5585585585585585 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 93, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + }, + "aya-original-human": { + "rows": 29, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 29, + "tp": 29, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 29, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 150, + "accuracy": 0.9666666666666667, + "macro_f1": 0.9680464778503994, + "macro_f1_present_labels": 0.9680464778503994, + "per_label": { + "AR": { + "support": 70, + "tp": 65, + "fp": 0, + "fn": 5, + "precision": 1.0, + "recall": 0.9285714285714286, + "f1": 0.9629629629629629 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 5, + "fn": 0, + "precision": 0.8888888888888888, + "recall": 1.0, + "f1": 0.9411764705882353 + } + }, + "confusion": [ + [ + 65, + 0, + 5 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "deu": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 186, + "accuracy": 0.7903225806451613, + "macro_f1": 0.8009399198931909, + "macro_f1_present_labels": 0.8009399198931909, + "per_label": { + "AR": { + "support": 82, + "tp": 43, + "fp": 0, + "fn": 39, + "precision": 1.0, + "recall": 0.524390243902439, + "f1": 0.688 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 36, + "fn": 0, + "precision": 0.5909090909090909, + "recall": 1.0, + "f1": 0.7428571428571429 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 3, + "fn": 0, + "precision": 0.9454545454545454, + "recall": 1.0, + "f1": 0.9719626168224299 + } + }, + "confusion": [ + [ + 43, + 36, + 3 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "eng": { + "rows": 5, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 5, + "tp": 5, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 5, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 150, + "accuracy": 0.7333333333333333, + "macro_f1": 0.751191952597492, + "macro_f1_present_labels": 0.751191952597492, + "per_label": { + "AR": { + "support": 70, + "tp": 30, + "fp": 0, + "fn": 40, + "precision": 1.0, + "recall": 0.42857142857142855, + "f1": 0.6 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 2, + "fn": 0, + "precision": 0.9523809523809523, + "recall": 1.0, + "f1": 0.975609756097561 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 38, + "fn": 0, + "precision": 0.5128205128205128, + "recall": 1.0, + "f1": 0.6779661016949152 + } + }, + "confusion": [ + [ + 30, + 2, + 38 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "fr": { + "rows": 150, + "accuracy": 0.76, + "macro_f1": 0.7811671087533156, + "macro_f1_present_labels": 0.7811671087533156, + "per_label": { + "AR": { + "support": 70, + "tp": 34, + "fp": 0, + "fn": 36, + "precision": 1.0, + "recall": 0.4857142857142857, + "f1": 0.6538461538461539 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 36, + "fn": 0, + "precision": 0.5263157894736842, + "recall": 1.0, + "f1": 0.6896551724137931 + } + }, + "confusion": [ + [ + 34, + 0, + 36 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "ja": { + "rows": 150, + "accuracy": 0.9333333333333333, + "macro_f1": 0.9373219373219372, + "macro_f1_present_labels": 0.9373219373219372, + "per_label": { + "AR": { + "support": 70, + "tp": 60, + "fp": 0, + "fn": 10, + "precision": 1.0, + "recall": 0.8571428571428571, + "f1": 0.9230769230769231 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 10, + "fn": 0, + "precision": 0.8, + "recall": 1.0, + "f1": 0.8888888888888888 + } + }, + "confusion": [ + [ + 60, + 0, + 10 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "spa": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 186, + "accuracy": 0.8440860215053764, + "macro_f1": 0.8545133878467212, + "macro_f1_present_labels": 0.8545133878467212, + "per_label": { + "AR": { + "support": 82, + "tp": 53, + "fp": 0, + "fn": 29, + "precision": 1.0, + "recall": 0.6463414634146342, + "f1": 0.7851851851851852 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 28, + "fn": 0, + "precision": 0.65, + "recall": 1.0, + "f1": 0.7878787878787878 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 1, + "fn": 0, + "precision": 0.9811320754716981, + "recall": 1.0, + "f1": 0.9904761904761905 + } + }, + "confusion": [ + [ + 53, + 28, + 1 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "zho": { + "rows": 3, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 3, + "tp": 3, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "short": { + "rows": 929, + "accuracy": 0.8417653390742734, + "macro_f1": 0.8469445749752597, + "macro_f1_present_labels": 0.8469445749752597, + "per_label": { + "AR": { + "support": 449, + "tp": 302, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.6726057906458798, + "f1": 0.8042609853528628 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 302, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "unspecified": { + "rows": 929, + "accuracy": 0.8417653390742734, + "macro_f1": 0.8469445749752597, + "macro_f1_present_labels": 0.8469445749752597, + "per_label": { + "AR": { + "support": 449, + "tp": 302, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.6726057906458798, + "f1": 0.8042609853528628 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 302, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + } + }, + "actual_tokens": { + "min": 9, + "max": 32768 + }, + "forward_batches_ms": [ + 7138.325362931937, + 20.877185044810176, + 19.67246795538813, + 19.299722043797374, + 19.489847007207572, + 19.313878961838782, + 18.220369005575776, + 18.069003941491246, + 17.690050997771323, + 18.120278022252023, + 18.159667029976845, + 18.83825403638184, + 18.094961997121572, + 18.554485985077918, + 21.646736073307693, + 20.772389019839466, + 18.628726014867425, + 17.715877038426697, + 18.23048701044172, + 18.000961048528552, + 18.17912806291133, + 18.317410023882985, + 18.020904972217977, + 17.96193898189813, + 17.93277298565954, + 57.140661985613406, + 18.297206959687173, + 18.85265507735312, + 17.86524197086692, + 18.02249299362302, + 18.041340983472764, + 18.24610901530832, + 17.97376002650708, + 18.182258936576545, + 17.512104008346796, + 18.063965952023864, + 17.891981988213956, + 18.07054504752159, + 20.899881958030164, + 20.305432961322367, + 17.980684991925955, + 19.004215020686388, + 18.027939018793404, + 18.247300991788507, + 17.516574007458985, + 19.29691794794053, + 18.164891051128507, + 18.128087976947427, + 18.290097010321915, + 17.779922927729785, + 19.383021048270166, + 18.370570964179933, + 18.288260092958808, + 20.001595956273377, + 19.974722061306238, + 19.60317906923592, + 18.46761896740645, + 25.879641994833946, + 21.32161194458604, + 67.82619096338749, + 66.8274499475956, + 67.0328710693866, + 67.87156604696065, + 66.34944898542017, + 65.6514149159193, + 64.8158120457083, + 69.92864399217069, + 67.61880894191563, + 61.89206801354885, + 69.69579495489597, + 65.14124805107713, + 72.03296897932887, + 71.39926496893167, + 76.04432699736208, + 74.31276992429048, + 70.53207501303405, + 72.93997798115015, + 242.4739560810849, + 237.631244934164, + 236.15267500281334, + 243.34610905498266, + 233.1851359922439, + 247.32551199849695, + 245.15614099800587, + 228.48063590936363, + 237.3392099980265, + 222.71971602458507, + 224.52398692257702, + 230.29644892085344, + 251.83466100133955, + 246.70857994351536, + 235.7939110370353, + 232.3068710975349, + 233.56033593881875, + 229.35457597486675, + 898.6606820253655, + 884.4740580534562, + 885.0953739602119, + 917.1203889418393, + 901.3165140058845, + 899.4212590623647, + 884.0509259607643, + 909.6224809763953, + 865.516652003862, + 862.0706489309669, + 865.1024099672213, + 843.7351549509913, + 889.8438949836418, + 837.8107619937509, + 879.9146349774674, + 820.4324760008603, + 907.9147179145366, + 908.1653489265591, + 3256.707448977977, + 3420.562503975816, + 3417.6524529466406, + 3594.7950570844114, + 3433.8663649978116, + 3413.9636090258136, + 3344.1693080822006, + 3134.1756300535053, + 3304.6368720242754, + 3353.6964000668377, + 3370.9448899608105, + 3342.783577973023, + 3258.953638956882, + 3131.3253020634875, + 3339.7319950163364, + 3427.004609024152, + 3479.9413479631767, + 3547.474607010372 + ], + "precision": { + "parameters": "float32", + "forward_autocast": "bfloat16", + "device_type": "cuda" + }, + "data_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + } + ] +} diff --git a/evaluation/migration-plan.json b/evaluation/migration-plan.json new file mode 100644 index 0000000000000000000000000000000000000000..60722316e925b6cb89021438832f6d7096e59a68 --- /dev/null +++ b/evaluation/migration-plan.json @@ -0,0 +1,19 @@ +{ + "candidate_step": 400, + "adapter_sha256": "5d2e619fe94be057a19931ad43122d68e07b34924ff487b4aab89346ff8f4e94", + "test_used": false, + "precision": "bfloat16 autocast with FP32 parameters", + "criteria": { + "short_macro_f1_max_drop": 0.005, + "long_source_macro_f1_max_drop": 0.01, + "each_long_length_max_correct_loss": 1, + "natural_aya_max_correct_loss": 1, + "academic_and_authored_source_macro_f1_max_drop": 0.01 + }, + "dev_files": { + "corpora/modality-v2/dev.jsonl": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "corpora/modality-v2-context-v2/dev.jsonl": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + }, + "source_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "target_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114" +} diff --git a/label_mapping.json b/label_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..ec64224bd1d6536f0b6cd34ef9200616b4f3f8bd --- /dev/null +++ b/label_mapping.json @@ -0,0 +1,22 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/lora/README.md b/lora/README.md new file mode 100644 index 0000000000000000000000000000000000000000..004b08e8ab29bd06ee56081927fe00460807a2ba --- /dev/null +++ b/lora/README.md @@ -0,0 +1,32 @@ +# Modality LoRA variant + +These adapter weights reproduce the same frozen task model as the merged weights +in the parent repository. The base is `llm-semantic-router/Vela-1.0-Encoder-307M` at +`5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114`. Use Transformers 4.57.6, PEFT 0.18.1 and Torch 2.10. +The adapter and task config are coupled: architecture, label IDs and pooling must +be loaded from this directory. In particular, Domain uses CLS pooling; loading +only the base configuration gives a different model. + +Download a pinned revision of this model repository, then run from its root: + +```bash +python lora/load_adapter.py --adapter lora --text "Explain cellular respiration." +python lora/verify_adapter.py --adapter lora --merged . --output adapter-verification.json +``` + +By default, the helper downloads the exact parent revision. `--base-directory` +can supply an already verified local snapshot of that same parent. It is the +caller's responsibility to verify a supplied local snapshot against the parent +weight hash in `adapter-qualification.json` in the repository root. + +The verification uses FP32 CPU SDPA on three padded two-row English/Chinese +fixtures, including Unicode text. It requires probability atol 2e-4 / rtol 1e-4 +and identical argmax labels. This checks short numerical parity, not held-out +quality, 32K execution, performance or all inference engines. The model card +contains separate task-quality and engine qualification evidence. + +`config.json`, tokenizer files and runtime mappings match the frozen merged model. +The load helper emits logits; PII entity decoding and a serving threshold are +separate operations. Maximum model capacity does not choose a device memory budget. +The root model license also applies to these adapter weights; bundled helper code +is Apache-2.0 under `reproducibility/LICENSE`. diff --git a/lora/adapter_config.json b/lora/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..51112cc9db6aa04cecca43ec804dffb1633d28e0 --- /dev/null +++ b/lora/adapter_config.json @@ -0,0 +1,58 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "llm-semantic-router/Vela-1.0-Encoder-307M", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 64, + "lora_bias": false, + "lora_dropout": 0.05, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "head", + "classifier", + "classifier", + "score", + "classifier", + "score", + "classifier", + "score", + "classifier", + "score", + "classifier", + "score", + "classifier", + "score" + ], + "peft_type": "LORA", + "peft_version": "0.18.1", + "qalora_group_size": 16, + "r": 32, + "rank_pattern": {}, + "revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "target_modules": [ + "attn.Wo", + "mlp.Wo", + "mlp.Wi", + "attn.Wqkv" + ], + "target_parameters": null, + "task_type": "SEQ_CLS", + "trainable_token_indices": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/lora/adapter_model.safetensors b/lora/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..828203479e3efdef3682fa9cc22462eb133815c5 --- /dev/null +++ b/lora/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00 +size 29427500 diff --git a/lora/config.json b/lora/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70f38b0a638eb4c674449d78ecc6f05e9ae9bd6e --- /dev/null +++ b/lora/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ModernBertForSequenceClassification" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 2, + "classifier_activation": "gelu", + "classifier_bias": false, + "classifier_dropout": 0.0, + "classifier_pooling": "mean", + "cls_token_id": 1, + "decoder_bias": true, + "deterministic_flash_attn": false, + "dtype": "float32", + "embedding_dropout": 0.0, + "eos_token_id": 1, + "global_attn_every_n_layers": 3, + "global_rope_theta": 160000, + "gradient_checkpointing": false, + "hidden_activation": "gelu", + "hidden_size": 768, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "initializer_cutoff_factor": 2.0, + "initializer_range": 0.02, + "intermediate_size": 1152, + "label2id": { + "AR": 0, + "BOTH": 2, + "DIFFUSION": 1 + }, + "layer_norm_eps": 1e-05, + "local_attention": 128, + "local_rope_theta": 160000, + "mask_token_id": 4, + "max_position_embeddings": 32768, + "mlp_bias": false, + "mlp_dropout": 0.0, + "model_type": "modernbert", + "norm_bias": false, + "norm_eps": 1e-05, + "num_attention_heads": 12, + "num_hidden_layers": 22, + "pad_token_id": 0, + "position_embedding_type": "sans_pos", + "problem_type": "single_label_classification", + "repad_logits_with_grad": false, + "sep_token_id": 1, + "sparse_pred_ignore_index": -100, + "sparse_prediction": false, + "transformers_version": "4.57.6", + "vocab_size": 256000 +} diff --git a/lora/label_mapping.json b/lora/label_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..ec64224bd1d6536f0b6cd34ef9200616b4f3f8bd --- /dev/null +++ b/lora/label_mapping.json @@ -0,0 +1,22 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/lora/load_adapter.py b/lora/load_adapter.py new file mode 100644 index 0000000000000000000000000000000000000000..9ef401ba8e64da34a9c2af15859b11a4365470a6 --- /dev/null +++ b/lora/load_adapter.py @@ -0,0 +1,71 @@ +"""Load a task LoRA using its explicit frozen task config and parent revision.""" + +import argparse +import json +from pathlib import Path + +import torch +from peft import PeftConfig, PeftModel +from transformers import ( + AutoConfig, + AutoModelForSequenceClassification, + AutoModelForTokenClassification, + AutoTokenizer, +) + + +def load_adapter(adapter, base_directory=None, device="cpu"): + adapter = Path(adapter) + parent = PeftConfig.from_pretrained(adapter) + if not parent.revision or len(parent.revision) != 40: + raise ValueError("Adapter must pin an immutable parent revision") + config = AutoConfig.from_pretrained(adapter) + architectures = config.architectures or [] + if architectures == ["ModernBertForTokenClassification"]: + factory = AutoModelForTokenClassification + elif architectures == ["ModernBertForSequenceClassification"]: + factory = AutoModelForSequenceClassification + if config.classifier_pooling not in {"cls", "mean"}: + raise ValueError("Unknown sequence pooling contract") + else: + raise ValueError("Unknown task architecture") + if set(map(int, config.id2label)) != set(range(config.num_labels)): + raise ValueError("Non-contiguous label IDs") + for index, label in config.id2label.items(): + if config.label2id[label] != int(index): + raise ValueError("Inconsistent task label mapping") + base = factory.from_pretrained( + base_directory or parent.base_model_name_or_path, + revision=None if base_directory else parent.revision, + config=config, + torch_dtype=torch.float32, + attn_implementation="sdpa", + ) + model = ( + PeftModel.from_pretrained(base, adapter, is_trainable=False).to(device).eval() + ) + tokenizer = AutoTokenizer.from_pretrained(adapter) + return model, tokenizer + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--adapter", type=Path, required=True) + parser.add_argument( + "--base-directory", + type=Path, + help="Optional already verified snapshot of the pinned parent", + ) + parser.add_argument("--text", required=True) + args = parser.parse_args() + model, tokenizer = load_adapter(args.adapter, args.base_directory) + inputs = tokenizer(args.text, return_tensors="pt", truncation=False) + if inputs["input_ids"].shape[1] > model.config.max_position_embeddings: + raise ValueError("Input exceeds the task model token budget") + with torch.inference_mode(): + logits = model(**inputs).logits.float() + print(json.dumps({"labels": model.config.id2label, "logits": logits.tolist()})) + + +if __name__ == "__main__": + main() diff --git a/lora/modality_mapping.json b/lora/modality_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..db14d98cd37b981d8a8d1b27770c3fed99021768 --- /dev/null +++ b/lora/modality_mapping.json @@ -0,0 +1,12 @@ +{ + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/lora/special_tokens_map.json b/lora/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..be4ad795f49075ac5e56d8108ef63be314a1e967 --- /dev/null +++ b/lora/special_tokens_map.json @@ -0,0 +1,55 @@ +{ + "additional_special_tokens": [ + "", + "" + ], + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "cls_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "sep_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/lora/tokenizer.json b/lora/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..3f81aca7d6f978a836be323226630e754367f1a0 --- /dev/null +++ b/lora/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f +size 34363343 diff --git a/lora/tokenizer_config.json b/lora/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8b7a2254239805140f894102e5bdc199186057b3 --- /dev/null +++ b/lora/tokenizer_config.json @@ -0,0 +1,2018 @@ +{ + "add_bos_token": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "5": { + "content": "<2mass>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "6": { + "content": "[@BOS@]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "7": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "8": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "9": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "10": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "11": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "12": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "13": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "14": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "15": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "16": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "17": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "18": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "19": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "20": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "21": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "22": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "23": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "24": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "25": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "26": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "27": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "28": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "29": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "30": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "31": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "32": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "33": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "34": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "35": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "36": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "37": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "38": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "39": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "40": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "41": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "42": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "43": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "44": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "45": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "46": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "47": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "48": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "49": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "50": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "51": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "52": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "53": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "54": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "55": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "56": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "57": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "58": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "59": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "60": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "61": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "62": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "63": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "64": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "65": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "66": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "67": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "68": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "69": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "70": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "71": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "72": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "73": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "74": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "75": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "76": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "77": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "78": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "79": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "80": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "81": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "82": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "83": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "84": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "85": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "86": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "87": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "88": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "89": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "90": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "91": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "92": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "93": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "94": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "95": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "96": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "97": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "98": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "99": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "100": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "101": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "102": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "103": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "104": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "105": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "106": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "107": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "108": { + "content": "\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "109": { + "content": "\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "110": { + "content": "\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "111": { + "content": "\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "112": { + "content": "\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "113": { + "content": "\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "114": { + "content": "\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "115": { + "content": "\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "116": { + "content": "\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "117": { + "content": "\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "118": { + "content": "\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "119": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "120": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "121": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "122": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "123": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "124": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "125": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "126": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "127": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "128": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "129": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "130": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "131": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "132": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "133": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "134": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "135": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "136": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "137": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "138": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "139": { + "content": "▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "140": { + "content": "▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "141": { + "content": "▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "142": { + "content": "▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "143": { + "content": "▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "144": { + "content": "▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "145": { + "content": "▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "146": { + "content": "▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "147": { + "content": "▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "148": { + "content": "▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "149": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "150": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "152": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "153": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "154": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "155": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "156": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "157": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "158": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "159": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "160": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "161": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "162": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "163": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "164": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "165": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "166": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "167": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "168": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "169": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "170": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "172": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "173": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "174": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "175": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "171": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "176": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "177": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "178": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "179": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "180": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "181": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "182": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "183": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "184": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "185": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "186": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "187": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "188": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "189": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "190": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "191": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "192": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "193": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "194": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "195": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "196": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "197": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "198": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "199": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "200": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "201": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "202": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "203": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "204": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "205": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "206": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "207": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "208": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "209": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "210": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "211": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "212": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "213": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "214": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "215": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "216": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255968": { + "content": "[toxicity=0]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255969": { + "content": "\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255970": { + "content": "\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255971": { + "content": "\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255972": { + "content": "\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255973": { + "content": "\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255974": { + "content": "\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255975": { + "content": "\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255976": { + "content": "\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255977": { + "content": "\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255978": { + "content": "\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255979": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255980": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255981": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255982": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255983": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255984": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255985": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255986": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255987": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255988": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255989": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255990": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255991": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255992": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255993": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255994": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255995": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255996": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255997": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255998": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255999": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "", + "" + ], + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 32768, + "pad_token": "", + "padding_side": "right", + "sep_token": "", + "spaces_between_special_tokens": false, + "tokenizer_class": "PreTrainedTokenizerFast", + "unk_token": "" +} diff --git a/lora/verify_adapter.py b/lora/verify_adapter.py new file mode 100644 index 0000000000000000000000000000000000000000..288869d02859050d5e371cf4450cc5d5c127e862 --- /dev/null +++ b/lora/verify_adapter.py @@ -0,0 +1,93 @@ +"""Reproduce the short CPU adapter/merged numerical contract; not a quality benchmark.""" + +import argparse +import hashlib +import json +from pathlib import Path + +import torch +from transformers import ( + AutoModelForSequenceClassification, + AutoModelForTokenClassification, +) +from load_adapter import load_adapter + +CASES = [ + ["Contact Ada Example at ada.example@example.org.", "请解释细胞呼吸如何产生能量。"], + [ + "Résumé — 王明在办公室阅读报告。 Contact test.user@example.org.", + "The committee discussed public history and economic policy.", + ], + ["A short request.", "A fictional setting for a story about a quiet garden. " * 9], +] + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--adapter", type=Path, required=True) + parser.add_argument("--merged", type=Path, required=True) + parser.add_argument("--base-directory", type=Path) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + if args.output.exists(): + raise ValueError("Refusing to overwrite evidence") + torch.set_num_threads(8) + model, tokenizer = load_adapter(args.adapter, args.base_directory) + token_task = model.config.architectures == ["ModernBertForTokenClassification"] + factory = ( + AutoModelForTokenClassification + if token_task + else AutoModelForSequenceClassification + ) + reference = factory.from_pretrained( + args.merged, torch_dtype=torch.float32, attn_implementation="sdpa" + ).eval() + results = [] + for texts in CASES: + inputs = tokenizer(texts, return_tensors="pt", padding=True, truncation=False) + with torch.inference_mode(): + a, b = model(**inputs).logits.float(), reference(**inputs).logits.float() + ap, bp = a.softmax(-1), b.softmax(-1) + flips = int((a.argmax(-1) != b.argmax(-1)).sum()) + passed = bool( + torch.isfinite(a).all() + and torch.isfinite(b).all() + and torch.allclose(ap, bp, atol=2e-4, rtol=1e-4) + and flips == 0 + ) + results.append( + { + "batch": len(texts), + "padded_tokens": inputs["input_ids"].shape[1], + "attention_mask_counts": inputs["attention_mask"].sum(-1).tolist(), + "input_sha256": hashlib.sha256( + json.dumps(texts, ensure_ascii=False).encode() + ).hexdigest(), + "maximum_absolute_logit_difference": float((a - b).abs().max()), + "maximum_absolute_probability_difference": float((ap - bp).abs().max()), + "argmax_differences": flips, + "passed": passed, + } + ) + args.output.write_text( + json.dumps( + { + "dtype": "float32", + "provider": "Torch CPU SDPA", + "probability_atol": 2e-4, + "probability_rtol": 1e-4, + "all_argmax_equal_required": True, + "quality_benchmark": False, + "long_context_qualification": False, + "results": results, + }, + indent=2, + ) + + "\n" + ) + if not all(row["passed"] for row in results): + raise RuntimeError("Adapter/merged numerical contract failed") + + +if __name__ == "__main__": + main() diff --git a/modality_mapping.json b/modality_mapping.json new file mode 100644 index 0000000000000000000000000000000000000000..db14d98cd37b981d8a8d1b27770c3fed99021768 --- /dev/null +++ b/modality_mapping.json @@ -0,0 +1,12 @@ +{ + "label_to_idx": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "idx_to_label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + } +} diff --git a/model.safetensors b/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a971a4304ea924185740773de7d9fd32bab6487e --- /dev/null +++ b/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7 +size 1230144500 diff --git a/onnx/model.onnx b/onnx/model.onnx new file mode 100644 index 0000000000000000000000000000000000000000..68937068e40815b0c1ef381ab4af047292c364e6 --- /dev/null +++ b/onnx/model.onnx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74a08c52ce47f522071fd71acee080c0a3a74ad12e3d5fb19ae1dfd244f1059c +size 555905 diff --git a/onnx/model.onnx.data b/onnx/model.onnx.data new file mode 100644 index 0000000000000000000000000000000000000000..316bb4ea931d7f9ebd701ed7384f110e25445a9e --- /dev/null +++ b/onnx/model.onnx.data @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0372b44dd9a9ae0d93f0ee0211ec27644dd2ad94351c8022e94c953bd14c399 +size 1230130176 diff --git a/publication-manifest.json b/publication-manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..abc23ece43356fa305a60d6bdb3c1c7fe1b774ee --- /dev/null +++ b/publication-manifest.json @@ -0,0 +1,154 @@ +{ + "model_id": "llm-semantic-router/Vela-1.0-Encoder-307M-Modality", + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "weight_sha256": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "parameter_count": 307532547, + "encoder_parameter_count": 306939648, + "classifier_pooling": "mean", + "weights_unchanged_after_final_evaluation": true, + "quality_protocol": "FP32 complete-input argmax; authored three-way contracts, natural AR requests, academic AR retention and long stress separately reported", + "native_and_onnx_artifacts": "Native FP32, frozen standalone LoRA, and qualified static-B1 FP32 blocked ONNX; failed experimental graphs excluded", + "model_files": { + "config.json": "4cc261fd20f7acf8eb78c7c81491a57eaff5cbf5dbe73ed738fc87af97e729f2", + "label_mapping.json": "63c62f883903b470c24ab5c760f593d0dbe4a01823376ee312dbf77fa9b7686a", + "modality_mapping.json": "40890ec869e786e4ca6510367bd1a6cdc2feedaf8564b345b24b38bbce7e0edd", + "model.safetensors": "d6fdd660a7bbecbb95e91956ce8f2f1c12b2424b2a0a42ca65bab2f858de07e7", + "special_tokens_map.json": "6e3204c5a4004719c185007a745d1940471a6fb02521cfa0a00306dbb6bc5903", + "tokenizer.json": "22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f", + "tokenizer_config.json": "74a259bb1a3811a7e3028adcd07a65765d866d5e66e0e883f0994ccfa67e8455" + }, + "support_files": { + "LICENSE": "f1e043aceb9aaca635570145a4246cba490f87a907715806ea020d05f45838a9", + "NOTICE.md": "35140a10f7fcbb826428daf9b6e767123ea58d740ac95b9d236f884cc86d668e", + "README.md": "5e93a2f8be652a59b23c0f3374a01558ad7fa08244668c7532d696771e07ad8b", + "evaluation/candidate-config.json": "4cc261fd20f7acf8eb78c7c81491a57eaff5cbf5dbe73ed738fc87af97e729f2", + "evaluation/candidate-label_mapping.json": "63c62f883903b470c24ab5c760f593d0dbe4a01823376ee312dbf77fa9b7686a", + "evaluation/candidate-lock.json": "34c6015eb91e0fd500371f9bad66cf28600533acb0c6197022eff3b3eeaa3b5f", + "evaluation/candidate-modality_mapping.json": "40890ec869e786e4ca6510367bd1a6cdc2feedaf8564b345b24b38bbce7e0edd", + "evaluation/comparison.json": "619fcba21a3eb603c28aab2445055b60c9d7a67f9cef5f963bf29983196ee457", + "evaluation/engines/blocked-storage-provenance.json": "ee4de9798b5277218618b453bc8bea3884c877462b5a6c6f86ca07ddfe3f82a3", + "evaluation/engines/vela-modality-blocked-b1-native-gpu-32k-v1.json": "315d6053bf4af6f4a5eaa68fcc4c06dfb938fd7d4250c74700fe92fa6647c467", + "evaluation/engines/vela-modality-blocked-b1-onnx-gpu-v1.json": "7264fa901a81c6078cd8cabb825d38b18a39746b19eff3ecc692813a256ec1c6", + "evaluation/engines/vela-modality-candle-cpu-v1.json": "f1cd9c3c58fd9af874157374a9c30d6ebda07175bd8a52e5fd4a675d7f801a97", + "evaluation/engines/vela-modality-standalone-engine-reproduction-v1.json": "6ccab48c6f0c6594c083f6fc12a71388128a57f72b7ccc9af2aac53f2a343abc", + "evaluation/final-baseline-fp32.json": "33cabd73d0ed8d8677f072a59c6470a2d00f5ff0ea32b966851ef41ede6e31c8", + "evaluation/final-baseline-fp32.predictions.jsonl": "2fe80dcebad0334c54c8d1bd064346d2e9e702a00fb01fba4b7186cd518d8c4e", + "evaluation/final-candidate-fp32.json": "2af3b6eb92e55d5e834f741dde44320c249800f302cf5eaff45094c78ea98006", + "evaluation/final-candidate-fp32.predictions.jsonl": "2f28bc2e3f7c9a6cd3898efe6210202fc75c4d265158210c1f148635cf8a131c", + "evaluation/final-evaluation-plan.json": "2dfa9c27d0f05236b9eb8bdb943c84007751407cdc4d8b73b8a4ff046538f4d0", + "evaluation/final-label-manifest.jsonl": "2bf2b4787248172f9df19fdc2203b20e5c83f1967f355922e5e9ba5025064fe0", + "evaluation/freeze-decision.json": "b768ec118e5f334d525712d7db215957e0ffdf4f517f462cbbef131e725d5e3b", + "evaluation/migration-adaptation-plan.json": "e80a24d2537fb99f767ca032b7ce010af587f01849b2d63332024ad094251463", + "evaluation/migration-decision.json": "7a90a2271692a92b358872e5854e5fca2b9809d32b8c30dbf4064bb150ea0776", + "evaluation/migration-dev-old.json": "97d807cce2873da2c3bb00ca9aab4b58d66a26f4f9ed06cbef5658118acd83b7", + "evaluation/migration-dev-vela.json": "ecaa7927e83815e289eb0434cde024991798da0e6f14e04ecd17cf5b0c352059", + "evaluation/migration-plan.json": "453ca8360bb857a207ba383701312f64fa43cb2ab67442b3d049f016ad9eebb0", + "reproducibility/LICENSE": "c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4", + "reproducibility/README.md": "07530eb8403cb6238e73f1d2cab784922d347236a6f1d8b2baa63f7dead29c54", + "reproducibility/bundle-checks.json": "80fe40424704431af730fdbf4d62734c65f9471d6ade580b7f8e9102abde8d12", + "reproducibility/contract-tests.txt": "5633287c6133137bb9c0bd8200d445deb1c14766cd532891a6c157bea19a92bc", + "reproducibility/corpora/domain-v1/contract.json": "95026e8dde09856c0c099106fed22579d97f2e30157e51cc4777af2771e43073", + "reproducibility/corpora/domain-v1/manifest.json": "2f5242918ef2ec6ea97aa9b87ea8dc0dbb21e06b10adf5ec2c490c345b2d63fb", + "reproducibility/corpora/modality-v1/contract.json": "87c8b1584d114e307901a3e1a203d93433cffbea0e62e3810825009a5f280a2a", + "reproducibility/corpora/modality-v1/manifest.json": "f202327c04bcbc28b0e8984ad00d8f33274cd744954c98bfa20f605bd2a9d833", + "reproducibility/corpora/modality-v2/contract.json": "87c8b1584d114e307901a3e1a203d93433cffbea0e62e3810825009a5f280a2a", + "reproducibility/corpora/modality-v2/manifest.json": "390faf835e37715d31a2a5160911c2e0e18c30a86c1be131321d33c5cf2a38e0", + "reproducibility/corpora/modality-v2-context-v2/contract.json": "87c8b1584d114e307901a3e1a203d93433cffbea0e62e3810825009a5f280a2a", + "reproducibility/corpora/modality-v2-context-v2/manifest.json": "43e2339c706d9bb921ffeadf37d855e1b1a37ec659146774fc57da7e99c91d85", + "reproducibility/data-file-hashes.json": "adde77781834a0802912b4a8a65264b15317cf2c34f506e8a621f0445ee3c0cd", + "reproducibility/data-reproduction.json": "96d5fd587d37ccc0c8d06d10bfa8b539228b8ea70f739dab65c2054ec17c76e5", + "reproducibility/environment.json": "cdffff561c770544bc6c3629baea14acc3cb9eb5f463d38d0ca7c3eae04dc7bc", + "reproducibility/evaluate_final.py": "c89ef86bc194618d832beaccf7d805fe15dd05bc65c702625f2d17ac3898083c", + "reproducibility/initialization-reproduction.json": "0611a8ea461086f3bbaa7df5e696811d639463b886e3a1b45301ed9e698e6bad", + "reproducibility/modality-source-files.json": "24c12b23d8877456516ec1ae317e9cc258eef8b1451bd29ba93e622a4fc96782", + "reproducibility/onnx/README.md": "6d4643f7e9803b7143ab555d89bd46c33136345bb69b28f7302618991536901b", + "reproducibility/onnx/export_classifier.py": "15837367a7ac3a6af308b34c6e72cf1a9a6d1f90ce3bab9a9fdec8d3c117d211", + "reproducibility/onnx/generate_engine_reference.py": "d5f9c09553e16aad0181a01f0793f7490993e2a6bf0992802780e62fd41f53af", + "reproducibility/onnx/onnx_artifacts.py": "324d3d4f1de13643e09d64e69e72a562fad9e0d30eada9f793d0aab64e2e56d0", + "reproducibility/onnx/qualify_onnx_fp32.py": "1bbc9da83314da564d0c0b1d43cb593954f28cb67b8a0fa16716288831d80733", + "reproducibility/onnx/requirements.txt": "1eda4cc7cb6ecfce45f2c646cc8c78ffaf08da08d310b8f1877e03fc9fcf5768", + "reproducibility/onnx/rewrite_blocked_attention.py": "cbb8d05dab9c257c4977d071fef11e0fba54b8d63c0fa7ada5f28608b13aec5d", + "reproducibility/onnx/rewrite_graph.py": "34672e6260b00a6c43298e55e9d082ee0e6c9e3d6102785f3e16bcfd3e56cfd8", + "reproducibility/onnx/stable_pooling.py": "8b67fdd8b11a975fe6a2a09e33970b7969c3f682f81f0e6cdb1ec534c90bb221", + "reproducibility/onnx/test_rewrite_blocked_attention.py": "7992368e87c4b3291157301bb907dbff273371d0e913ae37113a04d85c6c0336", + "reproducibility/prepare_data.py": "033d3b8095bcf50a90f6d12b25ef56a61202396ae487dea07d1977c068d4d302", + "reproducibility/release-lineage.json": "693e7564b755efb974b40b1bf72f873134ce3edc27effd08dd4a68781146e54a", + "reproducibility/requirements.txt": "fc5132424eb84042d86dca954ad29febd6fa3daec417a1e788bedd9526e41a1c", + "reproducibility/src/training/model_classifier/domain_classifier/README.md": "c6c2ab4cfe748dfcf643eb1f9b758dde897737996ccc75b8666b51a97ff1f5c5", + "reproducibility/src/training/model_classifier/domain_classifier/__init__.py": "c9744448e6d8dfc7555bf9dc5ca79ee913c5f45d7203e2a84316d5f020db8437", + "reproducibility/src/training/model_classifier/domain_classifier/prepare_data.py": "5a3bb37f15b6fcd0e9982224734bfd978a90fdc7a03b9025f40caa4ea9a7d6e4", + "reproducibility/src/training/model_classifier/modality_routing_classifier/extend_contracts.py": "b92671dc63baee1b94e32b7489587f2da5131f4a91a3ba04a40faa4fbb01cd1e", + "reproducibility/src/training/model_classifier/modality_routing_classifier/prepare_vela_data.py": "8e527a7f5190ae755b7c400a58caf0fa97a23d94a3711e668bd87225b9bd734b", + "reproducibility/src/training/model_classifier/sequence_repair/README.md": "258b6cab0851e54c33300d61df8ece8b6f346245cabaedc3623fdeeae4eb73a8", + "reproducibility/src/training/model_classifier/sequence_repair/__init__.py": "04cd2a4c037dd1cc2070d57bea979cacb76d4660313a190fdf612007cb33b88e", + "reproducibility/src/training/model_classifier/sequence_repair/complete_adapter.py": "f03c9c7743e1bfc051204c454d1d8230bfed8d4664dbcdeac48de3e7ea53f89f", + "reproducibility/src/training/model_classifier/sequence_repair/context.py": "8e51031478aca4a825a77d057b2a66a9d09d80e42fb66ccbec3c527f3a891303", + "reproducibility/src/training/model_classifier/sequence_repair/corpus.py": "ba3d4f91607db085e55ae1c76701b9b1319fea5ee7dece9851d4ee7375bca108", + "reproducibility/src/training/model_classifier/sequence_repair/data/README.md": "11960d9ca18463ccecd12bce5da30eba80b8bc3ab56b1ec5484320c6bedcd6ca", + "reproducibility/src/training/model_classifier/sequence_repair/data/aya-fact-reviewed-v2.json": "46127dfeaca954b0a38eef7a93188b736eec9fb1db5eaab833f2d89e51f98449", + "reproducibility/src/training/model_classifier/sequence_repair/data/aya-reviewed-labels-v1.json": "5ba7154d4ee94c15c2b6cd5fd494ea981fadc378048f0ad0436f99824c8745c3", + "reproducibility/src/training/model_classifier/sequence_repair/data/dolly-fact-reviewed-v2.json": "72aa07f8641b1fd369b365f54172fd5da2ceeb4a776bee8c8777a84e43e081cc", + "reproducibility/src/training/model_classifier/sequence_repair/data/source-files.json": "189c532e841346aa17191633f230be4eae0a69caa576c3385a1f2b75cbf9e346", + "reproducibility/src/training/model_classifier/sequence_repair/data/task-context-backgrounds.json": "b687f879fa4a5fa56d7fb7b9bef6a6f7b31fa7ad0cb3804365e3e45a68c97101", + "reproducibility/src/training/model_classifier/sequence_repair/data.py": "84f922c20a2f3b029f7d363cf7f96f98cd239669f7777de099c63cfa1a1ef199", + "reproducibility/src/training/model_classifier/sequence_repair/evaluate.py": "41617631838a1f72e33d402470a9bebd1ab0d7b42b80ac7e8c424a487068c562", + "reproducibility/src/training/model_classifier/sequence_repair/export.py": "334d5f27cbd953ef79f6a38b36c5ae9ba615eaf2449f46c16103f8800c04e662", + "reproducibility/src/training/model_classifier/sequence_repair/fetch_sources.py": "1abd30d33f456357553facdbde11116134835a087009c63c433faeccf8214638", + "reproducibility/src/training/model_classifier/sequence_repair/hydrate_annotations.py": "ea0c44b51afddbeb7dabf084db006eed545022ed819fa40682f4be846b207d62", + "reproducibility/src/training/model_classifier/sequence_repair/initialize.py": "d4e3a5ef2ee17759fac1a1bbfb95d165555dfd3d13c0334f38ae6464f2b82406", + "reproducibility/src/training/model_classifier/sequence_repair/model.py": "282b64def05cd9157062e4c1936848e30e0e7f18d5883605ad6d01e7f4828742", + "reproducibility/src/training/model_classifier/sequence_repair/prepare_context.py": "07d122e4025a25b30bf9c27a29881b4c18db5d908d135ee3f4df78115e54ffe7", + "reproducibility/src/training/model_classifier/sequence_repair/runtime_mapping.py": "ec0c553ef625001c09c53a27659cc5162629e303bdef6a93c6091767f49b7fd9", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_annotations.py": "5b331daaf2444b297752d5109dfe098ee05bb847c7fe8d350e826b208c72e34b", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_context.py": "b050b58fbba0bb62dc628d97703a071b5fde4a696a8031755e38f024ac4642a5", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_contracts.py": "3e1fd5e2fd7329c54b6a94a651050202666cf6e63a0555486e604c88a0a24d8f", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_recipe_sources.py": "2687d68f6b55021cbe81dbb35e90ded951918568e70fe46ec38cb55301ac62d4", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_runtime_mapping.py": "46e937fc5c43929c457a1d8709112da331fb557ec4c610e9ff17e0530757e927", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_sampling.py": "dc760c7459fb37ea007caf71e40232ad88854d36d4c20f1475c1f26b3d7ff5a9", + "reproducibility/src/training/model_classifier/sequence_repair/tests/test_token_budget.py": "abb0cee3bfaa1e19aadd7336c35d58caad7b1d6e92d2a0511c47dac86bea3262", + "reproducibility/src/training/model_classifier/sequence_repair/train.py": "2b6a160522dd6efd24893f6146bb641d7f1a8413bd690587b728c028e85d7e3e", + "reproducibility/summarize.py": "913832bf454e4c29c5a6c579d7999b564799a2b88e61499915c2a95cbb6e9a0b", + "reproducibility/train_stages.py": "87902e0e2d3f1729a32794179f0469134e3e740f1dad59d4faaaed15253f3249", + "reproducibility/training-receipts/modality-v1/run.json": "61232f37e3a58d7ac90798cf8ed75f184fcc8d2a7405b7449b09989ab78d6e58", + "reproducibility/training-receipts/modality-v1/selection.json": "cb56165aaa2c6db4d8ed10015a2b14f4ba5e977d1b720301d33d1b390e9d123b", + "reproducibility/training-receipts/modality-v2-diversity/run.json": "70d958965b08928bed3d54a982d29a2767a2cb9ba2bd4182c999081ae2ac7eba", + "reproducibility/training-receipts/modality-v2-diversity/selection.json": "137c71e36dc8629cbcca1665a4d0d8793e22490afc6f746a810aea2c7a1a63cb", + "reproducibility/training-receipts/modality-v3-context/run.json": "0104f13c1a43ec534edfec55661de81118b370c23d9eb40028153b47951940de", + "reproducibility/training-receipts/modality-v3-context/selection.json": "076c584384d8784253660d926e308489ec825a3055a1ebec0614b356b7b1fb8f", + "reproducibility/training-receipts/modality-v4-diverse-4k/run.json": "0f0f414b3bf84a3d166060f8772d52af92e646ea0335b0fb8dafedeec38e9200", + "reproducibility/training-receipts/modality-v4-diverse-4k/selection.json": "a9b4a16ee64ecb0f03f33c92b24bcceecf241e48ca72d6b478f8f8976455f70a", + "reproducibility/training-receipts/modality-v5-diverse-context/run.json": "d6d5959c097abed615e5a22770088056ec3a6502520ed1e4e3e0104ad022d7d7", + "reproducibility/training-receipts/modality-v5-diverse-context/selection.json": "a6d85b02bf9e5e1b07c7b8eff56ce00ee14e36ea0ade7ada4f742eb290ed7054", + "reproducibility/training-receipts/modality-v6-vela-base/dev-step-0.json": "589d9a69d3f81eb5e6624f36ed90b88197b167aba8ebbaea5656d7b200f6603c", + "reproducibility/training-receipts/modality-v6-vela-base/dev-step-200.json": "dbe84e85a80e91b1a7f307f2560531a7dc85db59b50017170602356b58171731", + "reproducibility/training-receipts/modality-v6-vela-base/run.json": "a6e0136fa8a08b0e03dd581518c2cda1d1a3468c5e32d84be0690888b350baf8", + "reproducibility/training-receipts/modality-v6-vela-base/selection.json": "c3c4f945b184eae3556c6931a13a23d22bac3155225cbee98e031121a3d27191" + }, + "onnx_variant": { + "precision": "float32", + "attention": "query-blocked, complete keys/values", + "static_batch_size": 1, + "qualified_max_sequence_length": 32768, + "custom_op_required": false, + "independent_task_quality": false + }, + "additional_files": { + "onnx/model.onnx": "74a08c52ce47f522071fd71acee080c0a3a74ad12e3d5fb19ae1dfd244f1059c", + "onnx/model.onnx.data": "f0372b44dd9a9ae0d93f0ee0211ec27644dd2ad94351c8022e94c953bd14c399", + "adapter-qualification.json": "1c3f3675e975a08f26685ff955263f2c17e6de624115b3a627ea90ca1b5f9c98", + "lora/README.md": "46387be05f1fa1bd1fd92c09f615bf8e3bd8d9b499089f785148c1c186f0bd8e", + "lora/adapter_config.json": "426d2ca8619487565d658930462e30a0d156641f0800c255f84b3610f44139d5", + "lora/adapter_model.safetensors": "52fec9bffae938380605bd1da242fcfded6c75964b89ce112de5206f14f4eb00", + "lora/config.json": "4cc261fd20f7acf8eb78c7c81491a57eaff5cbf5dbe73ed738fc87af97e729f2", + "lora/label_mapping.json": "63c62f883903b470c24ab5c760f593d0dbe4a01823376ee312dbf77fa9b7686a", + "lora/load_adapter.py": "759023bdc0bd563b422068c6b3dd28361a5c2ecdd3f4b8073ef931e040940298", + "lora/modality_mapping.json": "40890ec869e786e4ca6510367bd1a6cdc2feedaf8564b345b24b38bbce7e0edd", + "lora/special_tokens_map.json": "6e3204c5a4004719c185007a745d1940471a6fb02521cfa0a00306dbb6bc5903", + "lora/tokenizer.json": "22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f", + "lora/tokenizer_config.json": "74a259bb1a3811a7e3028adcd07a65765d866d5e66e0e883f0994ccfa67e8455", + "lora/verify_adapter.py": "ad9a6a32d8bc488dd56976936c4cccc217939a9c29256a069d4e9430c901b85f", + "standalone-verification.json": "1f1004e815211afa6ca714f169e311a54a6b27cc539de304f0d1064ebb6b1421", + "adapter-variant-manifest.json": "eeef8cc8f526a7a2c06778e9b8b0c8149670335f8c4be2f8638a4ac93ca59efc" + } +} diff --git a/reproducibility/LICENSE b/reproducibility/LICENSE new file mode 100644 index 0000000000000000000000000000000000000000..261eeb9e9f8b2b4b0d119366dda99c6fd7d35c64 --- /dev/null +++ b/reproducibility/LICENSE @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright [yyyy] [name of copyright owner] + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/reproducibility/README.md b/reproducibility/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9922d4d93af44ddac42cb54c59e93fcf9a33f48d --- /dev/null +++ b/reproducibility/README.md @@ -0,0 +1,114 @@ +# Reproduce Vela Modality text-output routing + +This text-only encoder classifies a requested output as `AR`, `DIFFUSION`, or +`BOTH`. It does not process images or audio. The released task starts from a +fresh adapter and head, then passes a development-only migration to the published +Vela Base. The old Modality weights are used only as a baseline. + +Use Python 3.12, an appropriate Torch 2.10 build, and `requirements.txt`. +Training uses FP32 parameters with BF16 autocast, SDPA and checkpointing; final +scoring uses FP32 throughout. Reproducing the recipe does not guarantee identical +weights across hardware, kernels or package versions. Run the commands below +from this directory. No training command automatically evaluates final data. + +## Reconstruct data and initialize + +```bash +python -m src.training.model_classifier.sequence_repair.fetch_sources \ + --manifest modality-source-files.json --output artifacts/sources +hf download llm-semantic-router/mmbert-32k-yarn \ + --revision 72a23a6640489471eb4ff7ad3ec5bc80af8a27de \ + --local-dir artifacts/base +hf download llm-semantic-router/Vela-1.0-Encoder-307M \ + --revision 5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114 \ + --local-dir artifacts/vela-base +python prepare_data.py --sources artifacts/sources \ + --tokenizer artifacts/base --output artifacts/data +python train_stages.py --stage initialize --artifacts artifacts +``` + +Source files and revisions are pinned for Global-MMLU, Aya, historical-source +screening and DiffusionDB. Only caption metadata are used from DiffusionDB; no +images are downloaded. Authored multilingual task contracts are combined with +captions. Text-answer academic questions and reviewed Aya requests provide AR +retention examples. They are not naturally balanced three-way modality labels. +Natural source requests are individually assistant-reviewed; this is not expert +human annotation. Data keep parent source groups together, and development/final +contract families are separated from training. Captions translated or reused +across variants do not count as independent examples. + +The diversity extension adds training-only contract families and neutral negative +contexts. Long data wrap a selected task payload in authored background prose, +retaining its source split and group. Train and evaluation backgrounds differ. +The 4K stage selects the 4K subset of the already frozen long data. All 14 data +files were reproduced byte-for-byte; their hashes are checked by the wrapper. +The public fresh initializer also reproduced the initial adapter weight hash. + +## Train the selected lineage + +```bash +python train_stages.py --stage modality-v1 --artifacts artifacts +python train_stages.py --stage modality-v2-diversity --artifacts artifacts +python train_stages.py --stage modality-v4-diverse-4k --artifacts artifacts +python train_stages.py --stage modality-v5-diverse-context --artifacts artifacts +python train_stages.py --stage modality-v6-vela-base --artifacts artifacts +python train_stages.py --stage export --artifacts artifacts +``` + +The original selected steps are 100, 300, 400, 400 and 200. Exact data hashes, +learning rates, sampling, seeds and selection metrics are in `training-receipts/`. +A rerun must select using its own development results. The omitted v3-context +trial is a diagnostic branch, not a parent of v4. The model retains standard +mean pooling throughout; complete task head and classifier weights are saved in +the adapter. + +The final 200-step stage uses the exact published Vela Base and preserves the +selected task adapter/head. Its direct-transfer baseline did not pass retention; +the bounded adaptation passes every predeclared short/source/length criterion. +The same source-macro F1 rule chooses step 200. It uses the explicit token-budget +microbatch option: the same pre-sampled global four examples are grouped to +reduce padding while preserving each example's CE weight. Dropout random +trajectories can differ across regroupings. The earlier stages retain their +original ungrouped execution path. + +The exporter preserves `ModernBertForSequenceClassification`, mean pooling, +`AR=0`, `DIFFUSION=1`, `BOTH=2`, and both generic and `modality_mapping.json` +runtime mappings. It locks config, tokenizer, mappings and weights together. + +## Independently evaluate the frozen final corpus + +```bash +hf download llm-semantic-router/mmbert32k-modality-router-merged \ + --revision 9f05f15889d32987d29bbe2e71bff945902ff553 \ + --local-dir artifacts/baseline +python evaluate_final.py --artifacts artifacts --model baseline +python evaluate_final.py --artifacts artifacts --model candidate +``` + +FP32 is the default. Both evaluations use identical gold, complete input without +truncation, fixed label IDs and argmax decisions. Existing outputs are not +overwritten. Final data are hash-checked against the freeze decision; there is +no final-based checkpoint or threshold selection. Report academic/AR retention, +natural AR requests, authored output contracts and long stress separately. +Long position/length variants are correlated stress cases, not a natural +long-document benchmark. Recorded forward time excludes tokenization, HTTP, +queueing and router execution. + +The package's contract tests are CPU-only and do not run training: + +```bash +python -m unittest discover \ + -s src/training/model_classifier/sequence_repair/tests -v +``` + +Recompute the source-separated results from the supplied text-free manifest: + +```bash +python summarize.py --manifest ../evaluation/final-label-manifest.jsonl \ + --baseline ../evaluation/final-baseline-fp32.predictions.jsonl \ + --candidate ../evaluation/final-candidate-fp32.predictions.jsonl \ + --contract corpora/modality-v2/contract.json --output comparison.json +``` + +All 33 supplied CPU contract tests passed, including the actual tiny-ModernBERT +gradient checks. The public log redacts its execution directory. diff --git a/reproducibility/bundle-checks.json b/reproducibility/bundle-checks.json new file mode 100644 index 0000000000000000000000000000000000000000..cdba47cb35dfac2e7516319f39d99af391bd3a96 --- /dev/null +++ b/reproducibility/bundle-checks.json @@ -0,0 +1,15 @@ +{ + "python_syntax": "all supplied Python modules parsed", + "markdown_links": "all relative links resolved", + "contract_tests": { + "tests": 33, + "passed": true, + "skipped": 0, + "original_log_sha256": "79a494a95cc00d6399518acf48943cf78c1e2df97636dbf379fdc60cd8cd43bc", + "public_log_path_redaction": true + }, + "data_reconstruction": "14 frozen data files reproduced byte-for-byte", + "initializer": "public fresh initializer weight hash matches original initialization", + "frozen_predictions": "all prediction/report bytes verified against their original SHA256; source-separated metrics recomputed", + "test_used_for_selection": false +} diff --git a/reproducibility/contract-tests.txt b/reproducibility/contract-tests.txt new file mode 100644 index 0000000000000000000000000000000000000000..129e44c3ede306abd82b04ca99e17ae3011c1a86 --- /dev/null +++ b/reproducibility/contract-tests.txt @@ -0,0 +1,48 @@ +test_aya_author_and_text_both_identify_the_reviewed_request (test_annotations.AnnotationTests.test_aya_author_and_text_both_identify_the_reviewed_request) ... ok +test_dolly_context_is_reconstructed_and_verified_not_discarded (test_annotations.AnnotationTests.test_dolly_context_is_reconstructed_and_verified_not_discarded) ... ok +test_estimated_search_matches_complete_paragraph_brute_force (test_context.ContextTests.test_estimated_search_matches_complete_paragraph_brute_force) ... ok +test_oversized_payload_is_rejected_without_truncation (test_context.ContextTests.test_oversized_payload_is_rejected_without_truncation) ... ok +test_payload_selection_preserves_distinct_source_groups_and_language (test_context.ContextTests.test_payload_selection_preserves_distinct_source_groups_and_language) ... ok +test_payload_survives_every_position_at_exact_attended_budget (test_context.ContextTests.test_payload_survives_every_position_at_exact_attended_budget) ... ok +test_accumulated_mean_cross_entropy_matches_full_batch_gradient (test_contracts.ContractTests.test_accumulated_mean_cross_entropy_matches_full_batch_gradient) ... /opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +/opt/amdgpu/share/libdrm/amdgpu.ids: No such file or directory +ok +test_conflicting_annotations_are_rejected (test_contracts.ContractTests.test_conflicting_annotations_are_rejected) ... ok +test_explicit_pooling_preserves_legacy_default_and_rejects_unknown_modes (test_contracts.ContractTests.test_explicit_pooling_preserves_legacy_default_and_rejects_unknown_modes) ... ok +test_label_ids_are_bijective_not_resorted (test_contracts.ContractTests.test_label_ids_are_bijective_not_resorted) ... ok +test_length_sampling_keeps_a_rare_long_bucket_visible (test_contracts.ContractTests.test_length_sampling_keeps_a_rare_long_bucket_visible) ... ok +test_length_selection_cannot_be_dominated_by_dense_bucket (test_contracts.ContractTests.test_length_selection_cannot_be_dominated_by_dense_bucket) ... ok +test_macro_f1_exposes_failure_on_minority_label (test_contracts.ContractTests.test_macro_f1_exposes_failure_on_minority_label) ... ok +test_present_source_selection_does_not_penalize_unobserved_labels (test_contracts.ContractTests.test_present_source_selection_does_not_penalize_unobserved_labels) ... ok +test_small_reviewed_source_remains_visible_with_length_and_label_balance (test_contracts.ContractTests.test_small_reviewed_source_remains_visible_with_length_and_label_balance) ... ok +test_source_selection_does_not_hide_a_reviewed_source_regression (test_contracts.ContractTests.test_source_selection_does_not_hide_a_reviewed_source_regression) ... ok +test_unicode_normalized_text_and_translation_groups_cannot_leak (test_contracts.ContractTests.test_unicode_normalized_text_and_translation_groups_cannot_leak) ... ok +test_unicode_separator_inside_json_string_is_not_a_record_boundary (test_contracts.ContractTests.test_unicode_separator_inside_json_string_is_not_a_record_boundary) ... ok +test_caption_recovery_keeps_embedded_colons_and_rejects_changed_source (test_recipe_sources.SourceContractTests.test_caption_recovery_keeps_embedded_colons_and_rejects_changed_source) ... ok +test_domain_question_contains_choices_but_not_answer_key (test_recipe_sources.SourceContractTests.test_domain_question_contains_choices_but_not_answer_key) ... ok +test_translated_duplicate_groups_are_unioned_transitively (test_recipe_sources.SourceContractTests.test_translated_duplicate_groups_are_unioned_transitively) ... ok +test_domain_uses_category_loader_schema_and_same_ids (test_runtime_mapping.RuntimeMappingTests.test_domain_uses_category_loader_schema_and_same_ids) ... ok +test_factcheck_polarity_is_preserved_in_both_directions (test_runtime_mapping.RuntimeMappingTests.test_factcheck_polarity_is_preserved_in_both_directions) ... ok +test_modality_cannot_reorder_native_label_indices (test_runtime_mapping.RuntimeMappingTests.test_modality_cannot_reorder_native_label_indices) ... ok +test_prompt_guard_and_dynamic_labels_use_router_layout (test_runtime_mapping.RuntimeMappingTests.test_prompt_guard_and_dynamic_labels_use_router_layout) ... ok +test_sparse_or_duplicate_ids_fail_before_writing (test_runtime_mapping.RuntimeMappingTests.test_sparse_or_duplicate_ids_fail_before_writing) ... ok +test_invalid_or_incomplete_source_weights_are_rejected (test_sampling.SourceSamplingTests.test_invalid_or_incomplete_source_weights_are_rejected) ... ok +test_missing_weights_preserve_legacy_rng_sequence (test_sampling.SourceSamplingTests.test_missing_weights_preserve_legacy_rng_sequence) ... ok +test_unique_coverage_keeps_unseen_rows_visible (test_sampling.SourceSamplingTests.test_unique_coverage_keeps_unseen_rows_visible) ... ok +test_weight_order_uses_source_occurrence_and_controls_frequency (test_sampling.SourceSamplingTests.test_weight_order_uses_source_occurrence_and_controls_frequency) ... ok +test_fp32_modernbert_gradient_matches_full_batch_with_dropout_disabled (test_token_budget.TokenBudgetTests.test_fp32_modernbert_gradient_matches_full_batch_with_dropout_disabled) ... artifacts/reference-venv/lib/python3.12/site-packages/torch/jit/_script.py:362: DeprecationWarning: `torch.jit.script_method` is deprecated. Please switch to `torch.compile` or `torch.export`. + warnings.warn( +ok +test_prefetch_preserves_sampling_and_rng_across_optimizer_steps (test_token_budget.TokenBudgetTests.test_prefetch_preserves_sampling_and_rng_across_optimizer_steps) ... ok +test_stable_groups_preserve_duplicate_draws_and_bound_padding (test_token_budget.TokenBudgetTests.test_stable_groups_preserve_duplicate_draws_and_bound_padding) ... ok + +---------------------------------------------------------------------- +Ran 33 tests in 4.012s + +OK diff --git a/reproducibility/corpora/domain-v1/contract.json b/reproducibility/corpora/domain-v1/contract.json new file mode 100644 index 0000000000000000000000000000000000000000..78c341377cbfa33b285eeb07657f1cdc03734d01 --- /dev/null +++ b/reproducibility/corpora/domain-v1/contract.json @@ -0,0 +1,35 @@ +{ + "label2id": { + "biology": 0, + "business": 1, + "chemistry": 2, + "computer science": 3, + "economics": 4, + "engineering": 5, + "health": 6, + "history": 7, + "law": 8, + "math": 9, + "other": 10, + "philosophy": 11, + "physics": 12, + "psychology": 13 + }, + "id2label": { + "0": "biology", + "1": "business", + "2": "chemistry", + "3": "computer science", + "4": "economics", + "5": "engineering", + "6": "health", + "7": "history", + "8": "law", + "9": "math", + "10": "other", + "11": "philosophy", + "12": "physics", + "13": "psychology" + }, + "problem_type": "single_label_classification" +} diff --git a/reproducibility/corpora/domain-v1/manifest.json b/reproducibility/corpora/domain-v1/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..2de8d727e5e099600e099389fdd2cc4c81d09e8b --- /dev/null +++ b/reproducibility/corpora/domain-v1/manifest.json @@ -0,0 +1,109 @@ +{ + "files": { + "train": { + "rows": 14932, + "sha256": "0aafba7b03e832966289dd87e885cf0188bdc9a3d906a79682070654f11b8705" + }, + "dev": { + "rows": 1988, + "sha256": "0e73503b193de3f9acd806ac7bed27b8fbd66790d8e2c4791744f2d2639047fc" + }, + "test": { + "rows": 2592, + "sha256": "c41d9f1d20d7280516640348d50f32526919e4315f0e4f134fd0484141a7cfcc" + } + }, + "provenance": { + "global_mmlu_revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "historical_sources_screened": 12353, + "ambiguous_label_rows_quarantined": 0, + "screening": "Exact normalized text or word unigram/bigram TF-IDF cosine >=0.75 moves entire translated question group into train", + "historical_exposure_not_guaranteed": true, + "subject_mapping": { + "biology": [ + "anatomy", + "college_biology", + "high_school_biology", + "medical_genetics", + "virology" + ], + "business": [ + "business_ethics", + "management", + "marketing", + "professional_accounting", + "public_relations" + ], + "chemistry": [ + "college_chemistry", + "high_school_chemistry" + ], + "computer science": [ + "college_computer_science", + "high_school_computer_science", + "computer_security", + "machine_learning" + ], + "economics": [ + "econometrics", + "high_school_macroeconomics", + "high_school_microeconomics" + ], + "engineering": [ + "electrical_engineering" + ], + "health": [ + "clinical_knowledge", + "college_medicine", + "professional_medicine", + "human_aging", + "human_sexuality", + "nutrition" + ], + "history": [ + "high_school_european_history", + "high_school_us_history", + "high_school_world_history", + "prehistory" + ], + "law": [ + "international_law", + "jurisprudence", + "professional_law" + ], + "math": [ + "abstract_algebra", + "college_mathematics", + "elementary_mathematics", + "high_school_mathematics", + "high_school_statistics" + ], + "philosophy": [ + "philosophy", + "moral_disputes", + "moral_scenarios", + "formal_logic", + "logical_fallacies" + ], + "physics": [ + "astronomy", + "college_physics", + "conceptual_physics", + "high_school_physics" + ], + "psychology": [ + "high_school_psychology", + "professional_psychology" + ] + }, + "unmapped_subject_policy": "other", + "annotations_sha256": "32d8c4cfd5b063cccb685889404de7099e62ecee633418bc8278f88cbe81b84d", + "source_group_rows": { + "global-mmlu-domain": 19326, + "aya-original-human": 213 + } + }, + "test_used_for_selection": false, + "split_policy": "Group-disjoint before sampling; exact NFKC/case/whitespace overlap rejected" +} diff --git a/reproducibility/corpora/modality-v1/contract.json b/reproducibility/corpora/modality-v1/contract.json new file mode 100644 index 0000000000000000000000000000000000000000..0804064ee7b4b9484ea914d4a4599e6dbd078637 --- /dev/null +++ b/reproducibility/corpora/modality-v1/contract.json @@ -0,0 +1,13 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "problem_type": "single_label_classification" +} diff --git a/reproducibility/corpora/modality-v1/manifest.json b/reproducibility/corpora/modality-v1/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..517ce7f2c08b1f86ae9efea49fe4a5ba905db66b --- /dev/null +++ b/reproducibility/corpora/modality-v1/manifest.json @@ -0,0 +1,37 @@ +{ + "files": { + "train": { + "rows": 4485, + "sha256": "b92180b7cfb98d6232805f64ffd6448d74e6a82c23dd2c4310a0f803541a5ab8" + }, + "dev": { + "rows": 929, + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + "test": { + "rows": 1339, + "sha256": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53" + } + }, + "provenance": { + "diffusiondb_revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "gallery_license": "CC0-1.0", + "other_source_licenses": [ + "Apache-2.0" + ], + "domain_parent_manifest_sha256": "2f5242918ef2ec6ea97aa9b87ea8dc0dbb21e06b10adf5ec2c490c345b2d63fb", + "annotation_roles": { + "authored-output-contract-with-gallery-caption": "Original six-language output instructions around source gallery captions; captions are not claimed to be translated or natural requests", + "academic-text-answer": "Academic question requests a text answer, not image generation", + "aya-original-human": "Human-authored source requests, independently assistant-reviewed task labels" + }, + "image_task_policy": "AR includes text, code, source SVG/Mermaid, describing existing images; DIFFUSION requests a new image; BOTH explicitly requests a new image and a separate textual explanation", + "caption_counts": { + "test": 60, + "train": 200, + "dev": 40 + } + }, + "test_used_for_selection": false, + "split_policy": "Group-disjoint before sampling; exact NFKC/case/whitespace overlap rejected" +} diff --git a/reproducibility/corpora/modality-v2-context-v2/contract.json b/reproducibility/corpora/modality-v2-context-v2/contract.json new file mode 100644 index 0000000000000000000000000000000000000000..0804064ee7b4b9484ea914d4a4599e6dbd078637 --- /dev/null +++ b/reproducibility/corpora/modality-v2-context-v2/contract.json @@ -0,0 +1,13 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "problem_type": "single_label_classification" +} diff --git a/reproducibility/corpora/modality-v2-context-v2/manifest.json b/reproducibility/corpora/modality-v2-context-v2/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..35a5ee5cf75a78be3720254cfbb6fc26b2d10477 --- /dev/null +++ b/reproducibility/corpora/modality-v2-context-v2/manifest.json @@ -0,0 +1,37 @@ +{ + "files": { + "train": { + "rows": 864, + "sha256": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0" + }, + "dev": { + "rows": 72, + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + }, + "test": { + "rows": 72, + "sha256": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c" + } + }, + "provenance": { + "parent_manifest_sha256": "390faf835e37715d31a2a5160911c2e0e18c30a86c1be131321d33c5cf2a38e0", + "backgrounds_sha256": "b687f879fa4a5fa56d7fb7b9bef6a6f7b31fa7ad0cb3804365e3e45a68c97101", + "tokenizer_sha256": "609d8f4c067cd3950f88594c5a802616cea245823836ef5848ee4fc40aab5b6f", + "train_payloads_per_label_language": 12, + "eval_payloads_per_label_language": 1, + "budgets": [ + 4096, + 8192, + 16384, + 32768 + ], + "languages": [ + "en", + "zh" + ], + "scope": "Authored repeated multi-paragraph context around frozen source requests, not natural long-document accuracy. Label is the explicitly marked requested task. Paired length/position variants retain the parent source group and partition.", + "test_used": false + }, + "test_used_for_selection": false, + "split_policy": "Group-disjoint before sampling; exact NFKC/case/whitespace overlap rejected" +} diff --git a/reproducibility/corpora/modality-v2/contract.json b/reproducibility/corpora/modality-v2/contract.json new file mode 100644 index 0000000000000000000000000000000000000000..0804064ee7b4b9484ea914d4a4599e6dbd078637 --- /dev/null +++ b/reproducibility/corpora/modality-v2/contract.json @@ -0,0 +1,13 @@ +{ + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "id2label": { + "0": "AR", + "1": "DIFFUSION", + "2": "BOTH" + }, + "problem_type": "single_label_classification" +} diff --git a/reproducibility/corpora/modality-v2/manifest.json b/reproducibility/corpora/modality-v2/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..ab4e4388407a705d8581b16273ec3a6c14f888f0 --- /dev/null +++ b/reproducibility/corpora/modality-v2/manifest.json @@ -0,0 +1,25 @@ +{ + "files": { + "train": { + "rows": 18885, + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + }, + "dev": { + "rows": 929, + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + "test": { + "rows": 1339, + "sha256": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53" + } + }, + "provenance": { + "parent_manifest_sha256": "f202327c04bcbc28b0e8984ad00d8f33274cd744954c98bfa20f605bd2a9d833", + "training_additions": 14400, + "scope": "Train-only authored output-contract variations on original train-only gallery source groups. Frozen dev/test requests and wording remain unchanged. Captions remain in their source language.", + "source_license": "CC0-1.0 for DiffusionDB captions", + "test_used": false + }, + "test_used_for_selection": false, + "split_policy": "Group-disjoint before sampling; exact NFKC/case/whitespace overlap rejected" +} diff --git a/reproducibility/data-file-hashes.json b/reproducibility/data-file-hashes.json new file mode 100644 index 0000000000000000000000000000000000000000..0527be3b8392334c71aaa310f9595e767e6ae949 --- /dev/null +++ b/reproducibility/data-file-hashes.json @@ -0,0 +1,16 @@ +{ + "domain-v1/train.jsonl": "0aafba7b03e832966289dd87e885cf0188bdc9a3d906a79682070654f11b8705", + "domain-v1/dev.jsonl": "0e73503b193de3f9acd806ac7bed27b8fbd66790d8e2c4791744f2d2639047fc", + "domain-v1/test.jsonl": "c41d9f1d20d7280516640348d50f32526919e4315f0e4f134fd0484141a7cfcc", + "modality-v1/train.jsonl": "b92180b7cfb98d6232805f64ffd6448d74e6a82c23dd2c4310a0f803541a5ab8", + "modality-v1/dev.jsonl": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "modality-v1/test.jsonl": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "modality-v2/train.jsonl": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1", + "modality-v2/dev.jsonl": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "modality-v2/test.jsonl": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "modality-v2-context-v2/train.jsonl": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0", + "modality-v2-context-v2/dev.jsonl": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85", + "modality-v2-context-v2/test.jsonl": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c", + "modality-v2-context-4k-v2/train.jsonl": "9c02fd4abfc08e29a3fc1f207157631f65f2f24ecaae5ea01b5d5f8b056f6ba3", + "modality-v2-context-4k-v2/dev.jsonl": "f99b3cb0e2608f385944d0bcd4023bc611f49e7bab02857e29436d5d71a9f263" +} diff --git a/reproducibility/data-reproduction.json b/reproducibility/data-reproduction.json new file mode 100644 index 0000000000000000000000000000000000000000..4a0e08b29bbafcadc4a8a9171a4def641e7d8789 --- /dev/null +++ b/reproducibility/data-reproduction.json @@ -0,0 +1,89 @@ +{ + "data_only": true, + "files": [ + { + "file": "domain-v1/train.jsonl", + "expected": "0aafba7b03e832966289dd87e885cf0188bdc9a3d906a79682070654f11b8705", + "actual": "0aafba7b03e832966289dd87e885cf0188bdc9a3d906a79682070654f11b8705", + "byte_equal": true + }, + { + "file": "domain-v1/dev.jsonl", + "expected": "0e73503b193de3f9acd806ac7bed27b8fbd66790d8e2c4791744f2d2639047fc", + "actual": "0e73503b193de3f9acd806ac7bed27b8fbd66790d8e2c4791744f2d2639047fc", + "byte_equal": true + }, + { + "file": "domain-v1/test.jsonl", + "expected": "c41d9f1d20d7280516640348d50f32526919e4315f0e4f134fd0484141a7cfcc", + "actual": "c41d9f1d20d7280516640348d50f32526919e4315f0e4f134fd0484141a7cfcc", + "byte_equal": true + }, + { + "file": "modality-v1/train.jsonl", + "expected": "b92180b7cfb98d6232805f64ffd6448d74e6a82c23dd2c4310a0f803541a5ab8", + "actual": "b92180b7cfb98d6232805f64ffd6448d74e6a82c23dd2c4310a0f803541a5ab8", + "byte_equal": true + }, + { + "file": "modality-v1/dev.jsonl", + "expected": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "actual": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "byte_equal": true + }, + { + "file": "modality-v1/test.jsonl", + "expected": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "actual": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "byte_equal": true + }, + { + "file": "modality-v2/train.jsonl", + "expected": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1", + "actual": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1", + "byte_equal": true + }, + { + "file": "modality-v2/dev.jsonl", + "expected": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "actual": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2", + "byte_equal": true + }, + { + "file": "modality-v2/test.jsonl", + "expected": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "actual": "20700706f7fc8876bf1b1235689d16c5ce4512fba35c70bc043abeb436b0cc53", + "byte_equal": true + }, + { + "file": "modality-v2-context-v2/train.jsonl", + "expected": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0", + "actual": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0", + "byte_equal": true + }, + { + "file": "modality-v2-context-v2/dev.jsonl", + "expected": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85", + "actual": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85", + "byte_equal": true + }, + { + "file": "modality-v2-context-v2/test.jsonl", + "expected": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c", + "actual": "fa8ec4fc2a6398ce0ef078bb9f4998e0bf120922afc4a77ff1fed4b2c17e1d4c", + "byte_equal": true + }, + { + "file": "modality-v2-context-4k-v2/train.jsonl", + "expected": "9c02fd4abfc08e29a3fc1f207157631f65f2f24ecaae5ea01b5d5f8b056f6ba3", + "actual": "9c02fd4abfc08e29a3fc1f207157631f65f2f24ecaae5ea01b5d5f8b056f6ba3", + "byte_equal": true + }, + { + "file": "modality-v2-context-4k-v2/dev.jsonl", + "expected": "f99b3cb0e2608f385944d0bcd4023bc611f49e7bab02857e29436d5d71a9f263", + "actual": "f99b3cb0e2608f385944d0bcd4023bc611f49e7bab02857e29436d5d71a9f263", + "byte_equal": true + } + ] +} diff --git a/reproducibility/environment.json b/reproducibility/environment.json new file mode 100644 index 0000000000000000000000000000000000000000..eaf5ed06bfbe0d96e8a1320285f8305c0419e7c1 --- /dev/null +++ b/reproducibility/environment.json @@ -0,0 +1,12 @@ +{ + "python": "3.12", + "torch": "2.10.0+rocm7.0", + "transformers": "4.57.6", + "peft": "0.18.1", + "attention": "sdpa", + "parameters": "float32", + "training_autocast": "bfloat16", + "device_details_omitted": true, + "source_commit": "e68ba24e9d754979cc4c9fd5be8c4f6ea465cb95", + "hardware_bitwise_reproducibility_guaranteed": false +} diff --git a/reproducibility/evaluate_final.py b/reproducibility/evaluate_final.py new file mode 100644 index 0000000000000000000000000000000000000000..1fadd0941b3a23f9ce0d96ba05660978fd8bfc3d --- /dev/null +++ b/reproducibility/evaluate_final.py @@ -0,0 +1,14 @@ +"""Evaluate fixed Modality weights on the sealed text-request final corpus.""" +import argparse,hashlib,json,subprocess,sys +from pathlib import Path + +def main(): + p=argparse.ArgumentParser();p.add_argument('--artifacts',type=Path,required=True);p.add_argument('--model',choices=['baseline','candidate'],required=True);p.add_argument('--dtype',choices=['float32','bfloat16'],default='float32');a=p.parse_args();r=a.artifacts + plan=json.loads(Path('../evaluation/freeze-decision.json').read_text());files=[] + for name,want in plan['final_data_sha256'].items(): + path=r/'data'/Path(name).relative_to('corpora') + if hashlib.sha256(path.read_bytes()).hexdigest()!=want:raise ValueError('Final data differ from freeze') + files.append(path) + model=r/('baseline' if a.model=='baseline' else 'frozen') + subprocess.run([sys.executable,'-m','src.training.model_classifier.sequence_repair.evaluate','--base',str(model),'--contract',str(r/'data/modality-v1/contract.json'),'--data',*map(str,files),'--output',str(r/f'final-{a.model}-{a.dtype}.json'),'--dtype',a.dtype,'--max-length','32768'],check=True) +if __name__=='__main__':main() diff --git a/reproducibility/initialization-reproduction.json b/reproducibility/initialization-reproduction.json new file mode 100644 index 0000000000000000000000000000000000000000..4b1226bce3a91a45f06bb7ff9859640770ab2ca3 --- /dev/null +++ b/reproducibility/initialization-reproduction.json @@ -0,0 +1,15 @@ +{ + "expected_adapter_sha256": "c11b45350cdac82aafa2943397f73ccbc6e31c9ee7806cb4038cf50333b3434d", + "actual": { + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "seed": 20260913, + "rank": 32, + "alpha": 64, + "dropout": 0.05, + "classifier_pooling": "mean", + "adapter_sha256": "c11b45350cdac82aafa2943397f73ccbc6e31c9ee7806cb4038cf50333b3434d", + "trainable_parameters": 7351299 + }, + "weights_byte_identical": true +} diff --git a/reproducibility/modality-source-files.json b/reproducibility/modality-source-files.json new file mode 100644 index 0000000000000000000000000000000000000000..a202ba7ec445b3f78e859e98d83fecfd11830b4a --- /dev/null +++ b/reproducibility/modality-source-files.json @@ -0,0 +1,162 @@ +[ + { + "repo": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "license": "Apache-2.0", + "file": "README.md", + "bytes": 13659, + "sha256": "3ac54f63eec84100c155571945da01c7c18cd8ec97ea94639e2b94a6860e8573" + }, + { + "repo": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "license": "Apache-2.0", + "file": "data/train-00000-of-00001.parquet", + "bytes": 137195800, + "sha256": "51baa85043b569ba117f41516b32d1b0e4e2647fb203a08933f98a556fb1fb06" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "README.md", + "bytes": 53310, + "sha256": "80e44287371de876f7b1dbe613494649ace41082e9a7caff8deb96b9b26421d3" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "en/dev-00000-of-00001.parquet", + "bytes": 90574, + "sha256": "86e985fb70b49130d751c7c5b255767348a8590793dc794ff0f34f1b8bd2d16f" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "en/test-00000-of-00001.parquet", + "bytes": 3910925, + "sha256": "5baec81ba97243728cc9557b2cba2f03aaeedf8232473db9f20cf995c6b0a1a9" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "zh/dev-00000-of-00001.parquet", + "bytes": 84714, + "sha256": "6c1dc5daf941264c1de302acf9088c7ed436d25c22e311d583271c81e7c4056e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "zh/test-00000-of-00001.parquet", + "bytes": 3618771, + "sha256": "e1ba0c287b650043b8686fa5c506d1f2718b098a7f88f724ade55a76f0062ae7" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "es/dev-00000-of-00001.parquet", + "bytes": 97926, + "sha256": "f4c4420525ef97f7a88da1bf6d8169c0faa70f49de6c779c04db94e8f278fac1" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "es/test-00000-of-00001.parquet", + "bytes": 4327396, + "sha256": "efbcc4ce1cf14f2f43f7b05be49a9a1153b2893c59d2f660c3bb58679801dccb" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "fr/dev-00000-of-00001.parquet", + "bytes": 100882, + "sha256": "bb3dd1fdd13509b0c8342af49470bbceca9cae2bb32f6e3f589de93aa4c2a82e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "fr/test-00000-of-00001.parquet", + "bytes": 4550854, + "sha256": "e602201459c29c5dbacf9a3d80cfe2524496f8e0c0b51b6326ef4205bad3405e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "de/dev-00000-of-00001.parquet", + "bytes": 100678, + "sha256": "242f31aca9cfdcc7c87f6a77eef6ced0c7967153400898e10dc264a196116d76" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "de/test-00000-of-00001.parquet", + "bytes": 4543635, + "sha256": "b58a0d5994e756dabef96eb804693be153953052276d3e0b7d57bb2f2798854a" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "ja/dev-00000-of-00001.parquet", + "bytes": 98924, + "sha256": "63c3fd6d20d016e7dc0df1d9b830ab5d93f9150450a24ec992a4662f2a80fefc" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "Apache-2.0", + "file": "ja/test-00000-of-00001.parquet", + "bytes": 4363616, + "sha256": "b56ad153180d4368757891796dfad0e1a76ebeb3f7b2500b1ac63789d4e4dfe7" + }, + { + "repo": "TIGER-Lab/MMLU-Pro", + "revision": "b189ec765aa7ed75c8acfea42df31fdae71f97be", + "file": "data/test-00000-of-00001.parquet", + "sha256": "0e24a191921c2f453518a537a8b2117bd137e7714d4ef1565e9ba06c1ecb9ad8", + "role": "historical exposure screening", + "license": "MIT" + }, + { + "repo": "TIGER-Lab/MMLU-Pro", + "revision": "b189ec765aa7ed75c8acfea42df31fdae71f97be", + "file": "data/validation-00000-of-00001.parquet", + "sha256": "139423c23722e480c807ac4a191409a710cfce4eba744c1d641cf88e730e2078", + "role": "historical exposure screening", + "license": "MIT" + }, + { + "repo": "LLM-Semantic-Router/category-classifier-supplement", + "revision": "c51aed0d3a83a548270e835187cee30615188e60", + "file": "data/train-00000-of-00001.parquet", + "sha256": "bf5c5e52f9c58f25d0e2a9bc7d23e401db20fde8fac6dece364a6820ed44b7d8", + "role": "historical exposure screening", + "license": "Apache-2.0" + }, + { + "repo": "poloclub/diffusiondb", + "revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "license": "CC0-1.0", + "file": "README.md", + "bytes": 24993, + "sha256": "d4e51a582f661d61647fbd74c7974b8dc1b59a47fed20aa5a9597100875dd38e" + }, + { + "repo": "poloclub/diffusiondb", + "revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "license": "CC0-1.0", + "file": "metadata.parquet", + "bytes": 194548652, + "sha256": "eecd341187bc91c07f5994ad0660d40228ea025616fd57a509bef8323677c68f" + } +] diff --git a/reproducibility/onnx/README.md b/reproducibility/onnx/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae8d3a75dc2038d5670f6bd617f6999814f0abb3 --- /dev/null +++ b/reproducibility/onnx/README.md @@ -0,0 +1,46 @@ +# Reproduce FP32 export and GPU qualification + +These scripts preserve the frozen weights and task architecture. Use the +recorded Transformers 4.57.6 / Torch 2.10 environment, `onnx`, and a compatible +ONNX Runtime installation. CPU graph tests ran with ONNX Runtime 1.24.2; the +recorded AMD numerical qualification used its ROCm provider in version 1.22.1. +Select a Torch/ONNX Runtime build suitable for your hardware. Installing the +ordinary CPU-only ONNX Runtime wheel does not add a GPU provider. + +Install a hardware-appropriate Torch 2.10 build first (the measured exporter +used `2.10.0+rocm7.0`), then run `python -m pip install -r requirements.txt`. +The pinned requirements include `onnxscript`, which the Dynamo exporter requires. +For GPU qualification use a separate environment with the tested ROCm-enabled +ONNX Runtime 1.22.1; do not overwrite that provider build with the CPU wheel. + +Run from this directory. Download the complete model into `artifacts/model` +first, and preserve the native weight SHA256 from the model card. + +```bash +python export_classifier.py --model artifacts/model --output artifacts/portable --dtype float32 +python rewrite_blocked_attention.py artifacts/portable/model.onnx artifacts/blocked/model.onnx --batch-size 1 +python generate_engine_reference.py --model artifacts/model --output artifacts/reference.npz --task sequence --device cuda --lengths 2 63 64 65 127 128 129 512 4096 8192 16384 32768 +python qualify_onnx_fp32.py --model artifacts/model --reference artifacts/reference.npz --graph artifacts/blocked/model.onnx --batch-size 1 --sequence --output artifacts/qualification.json +python -m unittest test_rewrite_blocked_attention +``` + +These commands generate new deterministic numerical probes for independent +revalidation. Historical reports bind their original reference files by SHA256; +the new NPZ bytes and measured rounding differences need not be identical. + +The explicit long-reference command needs memory for FP32 attention at the +selected lengths and batches. Start with the default short lengths on a smaller +device. `cuda` is also the PyTorch device name for supported ROCm builds. +Reference generation includes padded B2 inputs for reproducible diagnostic +comparisons; the released blocked graph accepts only B1. Qualification checks +both static input dimensions and verifies that B2 is rejected. It compares every +class probability with FP32 HF (atol 2e-4, rtol 1e-4), requires identical argmax +labels, and checks that loop attention MatMul/Softmax execute on ROCm. It does +not infer task accuracy or an isolated performance benchmark from these probes. + +The ordinary FP32 export and blocked graph are distinct variants. Keep each +graph with its adjacent external data file. The released `onnx/model.onnx` is +the qualified static-B1 blocked variant; do not substitute experimental dynamic +or half-precision files without repeating qualification. The native HF model +has its own batching contract. These commands do not alter training or select a +checkpoint using final outcomes. diff --git a/reproducibility/onnx/export_classifier.py b/reproducibility/onnx/export_classifier.py new file mode 100644 index 0000000000000000000000000000000000000000..c2d8d4c5acb207ee12c758ceeae0a4291e0fd2da --- /dev/null +++ b/reproducibility/onnx/export_classifier.py @@ -0,0 +1,324 @@ +#!/usr/bin/env python3 +"""Export a local, merged ModernBERT task checkpoint and verify dynamic logits. + +The portable FP32 graph is model.onnx; FP16 is model_sdpa_fp16.onnx. +CK Flash Attention conversion and GPU parity are separate, explicit steps. +An export receipt proves numerical agreement, not task quality at that length. +""" + +import argparse +import json +from pathlib import Path + +import numpy as np +import onnx +import onnxruntime as ort +import torch +import transformers +from onnx_artifacts import external_data_sha256, sha256, strip_debug_annotations +from transformers import ( + AutoConfig, + AutoModelForSequenceClassification, + AutoModelForTokenClassification, + AutoTokenizer, +) + +MIN_VALIDATION_TOKENS = 2 + + +class ClassifierLogits(torch.nn.Module): + """Keep pooling and the complete task head in FP32, including FP16 exports.""" + + def __init__(self, model, token_classification): + super().__init__() + self.model = model + self.token_classification = token_classification + self.model.head.float() + self.model.classifier.float() + if not token_classification and model.config.classifier_pooling not in { + "mean", + "cls", + }: + raise ValueError("Only mean or cls sequence pooling is supported") + + def forward(self, input_ids, attention_mask): + hidden = self.model.model( + input_ids=input_ids, attention_mask=attention_mask + ).last_hidden_state + if self.token_classification: + return self.model.classifier( + self.model.drop(self.model.head(hidden.float())) + ) + if self.model.config.classifier_pooling == "cls": + pooled = hidden[:, 0].float() + else: + mask = attention_mask.unsqueeze(-1).float() + pooled = (hidden.float() * mask).sum(1) / mask.sum(1) + return self.model.classifier(self.model.drop(self.model.head(pooled))) + + +def make_input(seed_ids, length, batch, pad_token_id): + repeated = (seed_ids * (length // len(seed_ids) + 1))[:length] + ids = torch.tensor(repeated, dtype=torch.long).repeat(batch, 1) + mask = torch.ones_like(ids) + if batch > 1: + pad = max(1, length // 3) + mask[0, -pad:] = 0 + ids[0, -pad:] = pad_token_id + return ids, mask + + +def verify_output_source(output, source_artifacts): + """A directory's precision variants must describe one native checkpoint.""" + for path in output.glob("export-*.json"): + prior = json.loads(path.read_text()) + if prior.get("source_artifacts") != source_artifacts: + raise ValueError( + "Export directory contains a different or unpinned source; " + "use a fresh directory" + ) + + +def task_probabilities(logits, multi_label): + values = torch.from_numpy(logits).float() + return (values.sigmoid() if multi_label else values.softmax(-1)).numpy() + + +def verify_task_outputs(actual, expected, token_task, multi_label, dtype): + """Check task semantics while retaining raw-logit errors in the receipt.""" + scores = task_probabilities(actual, multi_label) + reference_scores = task_probabilities(expected, multi_label) + if dtype == "float32" and not token_task: + np.testing.assert_allclose(actual, expected, atol=2e-4, rtol=1e-4) + # Token consumers use the complete softmax vector and BIO argmax. Dormant + # class logits can amplify FP32 kernel rounding without changing either; + # raw-logit error is diagnostic, not a token-decision tolerance. + np.testing.assert_allclose( + scores, + reference_scores, + atol=3e-3 if dtype == "float16" else 2e-4, + rtol=1e-3 if dtype == "float16" else 1e-4, + ) + if token_task: + np.testing.assert_array_equal(actual.argmax(-1), expected.argmax(-1)) + return scores, reference_scores + + +def verify_graph( + reference, graph_path, seed_ids, tokenizer, lengths, token_task, multi_label, dtype +): + options = ort.SessionOptions() + options.intra_op_num_threads = 8 + options.inter_op_num_threads = 1 + session = ort.InferenceSession( + str(graph_path), options, providers=["CPUExecutionProvider"] + ) + results = [] + for length in lengths: + for batch in (1, 2): + ids, mask = make_input(seed_ids, length, batch, tokenizer.pad_token_id) + with torch.inference_mode(): + expected = reference(ids, mask).float().numpy() + actual = session.run( + None, {"input_ids": ids.numpy(), "attention_mask": mask.numpy()} + )[0].astype(np.float32) + if actual.shape != expected.shape or not np.isfinite(actual).all(): + raise ValueError("Graph returned invalid shape or nonfinite logits") + # A padded token's query output has no task meaning. Its key is still + # masked, so agreement on valid tokens verifies padding isolation. + valid = mask.numpy().astype(bool) if token_task else slice(None) + delta = np.abs(actual[valid] - expected[valid]) + # Quantization error is measured against full-precision native + # inference, never against platform-dependent CPU half kernels. + scores, reference_scores = verify_task_outputs( + actual[valid], expected[valid], token_task, multi_label, dtype + ) + results.append( + { + "batch": batch, + "sequence_length": length, + "valid_tokens": mask.sum(1).tolist(), + "output_shape": list(actual.shape), + "max_abs_logit_error": float(delta.max()), + "max_abs_probability_error": float( + np.abs(scores - reference_scores).max() + ), + "argmax_agreement": float( + (actual[valid].argmax(-1) == expected[valid].argmax(-1)).mean() + ), + } + ) + return results + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument( + "--model", type=Path, required=True, help="Immutable local merged checkpoint" + ) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument( + "--verify-only", + action="store_true", + help="Recheck an existing export against the checkpoint", + ) + parser.add_argument("--dtype", choices=("float32", "float16"), default="float32") + parser.add_argument( + "--validation-lengths", + type=int, + nargs="+", + default=[2, 63, 64, 65, 127, 128, 129, 512], + ) + args = parser.parse_args() + if not args.model.is_dir() or args.model.resolve() == args.output.resolve(): + raise ValueError("Use a local checkpoint and a separate export directory") + torch.set_num_threads(8) + source_artifacts = { + path.name: sha256(path) + for path in sorted(args.model.iterdir()) + if path.is_file() and path.suffix in {".json", ".safetensors"} + } + verify_output_source(args.output, source_artifacts) + config = AutoConfig.from_pretrained(args.model, local_files_only=True) + if config.model_type != "modernbert": + raise ValueError("This exporter only supports ModernBERT task heads") + architectures = config.architectures or [] + token_task = architectures == ["ModernBertForTokenClassification"] + if not token_task and architectures != ["ModernBertForSequenceClassification"]: + raise ValueError(f"A merged task checkpoint is required, got {architectures}") + if set(config.id2label) != set(range(config.num_labels)): + raise ValueError("id2label must exactly cover the task head") + if len(set(config.id2label.values())) != config.num_labels: + raise ValueError("Task labels must be unique") + if any( + length < MIN_VALIDATION_TOKENS or length > config.max_position_embeddings + for length in args.validation_lengths + ): + raise ValueError("Validation lengths must fit the model's context") + kind = ( + AutoModelForTokenClassification + if token_task + else AutoModelForSequenceClassification + ) + model, loading = kind.from_pretrained( + args.model, + local_files_only=True, + torch_dtype=torch.float32, + attn_implementation="sdpa", + reference_compile=False, + output_loading_info=True, + ) + if any( + loading.get(key) + for key in ("missing_keys", "unexpected_keys", "mismatched_keys", "error_msgs") + ): + raise ValueError(f"Incomplete task checkpoint: {loading}") + model.model.to(getattr(torch, args.dtype)) + wrapper = ClassifierLogits(model.eval(), token_task).eval() + reference = wrapper + if args.dtype == "float16": + reference = ClassifierLogits( + kind.from_pretrained( + args.model, + local_files_only=True, + torch_dtype=torch.float32, + attn_implementation="sdpa", + reference_compile=False, + ).eval(), + token_task, + ).eval() + tokenizer = AutoTokenizer.from_pretrained(args.model, local_files_only=True) + seed_ids = tokenizer.encode( + "Example context. 中文测试。 Contact alice@example.org." + ) + ids, mask = make_input(seed_ids, 64, 2, tokenizer.pad_token_id) + args.output.mkdir(parents=True, exist_ok=True) + path = args.output / ( + "model.onnx" if args.dtype == "float32" else "model_sdpa_fp16.onnx" + ) + if args.verify_only and not path.is_file(): + raise FileNotFoundError("No exported graph to verify") + if path.exists() and not args.verify_only: + raise FileExistsError( + "Use a fresh export directory to preserve earlier evidence" + ) + batch = torch.export.Dim("batch", min=1, max=16) + sequence = torch.export.Dim("sequence", min=2, max=config.max_position_embeddings) + if not args.verify_only: + with torch.inference_mode(): + torch.onnx.export( + wrapper, + (ids, mask), + str(path), + input_names=["input_ids", "attention_mask"], + output_names=["logits"], + opset_version=18, + dynamo=True, + external_data=True, + dynamic_shapes=({0: batch, 1: sequence}, {0: batch, 1: sequence}), + optimize=True, + ) + graph = onnx.load(path, load_external_data=False) + external_files = external_data_sha256(graph, path) + strip_debug_annotations(graph) + onnx.save(graph, path) + onnx.checker.check_model(str(path)) + expected_rank = 3 if token_task else 2 + if len(graph.graph.output[0].type.tensor_type.shape.dim) != expected_rank: + raise ValueError("Exported graph has the wrong task rank") + results = verify_graph( + reference, + path, + seed_ids, + tokenizer, + args.validation_lengths, + token_task, + config.problem_type == "multi_label_classification", + args.dtype, + ) + model.config._name_or_path = "" + model.config.save_pretrained(args.output) + tokenizer.save_pretrained(args.output) + receipt = { + "task": "token-classification" if token_task else "text-classification", + "dtype": args.dtype, + "head_dtype": "float32", + "pooling_accumulation_dtype": "float32", + "reference_dtype": "float32", + "output_validation": ( + "all valid-token probabilities and exact BIO argmax; raw logits reported" + if token_task + else "all class probabilities; raw logits also checked for FP32" + ), + "id2label": config.id2label, + "max_position_embeddings": config.max_position_embeddings, + "versions": { + "torch": torch.__version__, + "transformers": transformers.__version__, + "onnx": onnx.__version__, + }, + "source_weights": { + p.name: sha256(p) for p in sorted(args.model.glob("*.safetensors")) + }, + "source_artifacts": source_artifacts, + "exported_files": { + p.name: sha256(p) + for p in sorted(args.output.iterdir()) + if p.is_file() and not p.match("export-*.json") + }, + "external_data": external_files, + "cpu_dynamic_parity": results, + "scope": "Numerical parity only; task quality and long-context GPU behavior require separate evaluations.", + } + (args.output / f"export-{args.dtype}.json").write_text( + json.dumps(receipt, indent=2) + "\n" + ) + print( + json.dumps( + {"task": receipt["task"], "dtype": args.dtype, "parity_cases": len(results)} + ) + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/onnx/generate_engine_reference.py b/reproducibility/onnx/generate_engine_reference.py new file mode 100644 index 0000000000000000000000000000000000000000..0558c82797e9b3e2e1c3bfc8322638ab52d63f97 --- /dev/null +++ b/reproducibility/onnx/generate_engine_reference.py @@ -0,0 +1,69 @@ +"""Generate deterministic, fixed-weight FP32 references for ONNX qualification.""" + +import argparse +import hashlib +import json +from pathlib import Path + +import numpy as np +import torch +from transformers import ( + AutoModelForSequenceClassification, + AutoModelForTokenClassification, + AutoTokenizer, +) + +from export_classifier import ClassifierLogits, make_input + + +def sha256(path): + value = hashlib.sha256() + with path.open('rb') as stream: + for chunk in iter(lambda: stream.read(8 << 20), b''): + value.update(chunk) + return value.hexdigest() + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument('--model', type=Path, required=True) + parser.add_argument('--output', type=Path, required=True) + parser.add_argument('--task', choices=['token', 'sequence'], required=True) + parser.add_argument('--device', choices=['cpu', 'cuda'], required=True) + parser.add_argument('--lengths', type=int, nargs='+', default=[2, 63, 64, 65, 127, 128, 129, 512]) + args = parser.parse_args() + if args.output.exists() or args.output.suffix != '.npz': + raise ValueError('Use a new .npz reference file') + config = json.loads((args.model/'config.json').read_text()) + if any(not 2 <= value <= config['max_position_embeddings'] for value in args.lengths): + raise ValueError('Every probe length must fit the declared model context') + torch.set_num_threads(8) + factory = AutoModelForTokenClassification if args.task == 'token' else AutoModelForSequenceClassification + model = factory.from_pretrained(args.model, local_files_only=True, torch_dtype=torch.float32, + attn_implementation='sdpa', reference_compile=False).to(args.device).eval() + wrapper = ClassifierLogits(model, args.task == 'token').eval() + tokenizer = AutoTokenizer.from_pretrained(args.model, local_files_only=True) + seed = tokenizer.encode('Example context. 中文测试。 Contact alice@example.org.') + weight = sha256(args.model/'model.safetensors') + values = {'weights_sha256': np.array(weight), 'metadata': np.array(json.dumps({ + 'weights_sha256': weight, 'config_sha256': sha256(args.model/'config.json'), + 'task': args.task, 'torch_version': torch.__version__, 'device': args.device, + }))} + cases = [] + for length in args.lengths: + for batch in (1, 2): + ids, mask = make_input(seed, length, batch, tokenizer.pad_token_id) + with torch.inference_mode(): + logits = wrapper(ids.to(args.device), mask.to(args.device)).float().cpu().numpy() + if not np.isfinite(logits).all(): + raise ValueError(f'Nonfinite reference at batch {batch}, length {length}') + key = f'b{batch}s{length}' + values[key+'_ids'], values[key+'_mask'], values[key+'_logits'] = ids.numpy(), mask.numpy(), logits + cases.append(key) + print(key, flush=True) + values['cases'] = np.array(cases) + np.savez(args.output, **values) + + +if __name__ == '__main__': + main() diff --git a/reproducibility/onnx/onnx_artifacts.py b/reproducibility/onnx/onnx_artifacts.py new file mode 100644 index 0000000000000000000000000000000000000000..04951a846b41acbe24b174361f132c4cb78eb806 --- /dev/null +++ b/reproducibility/onnx/onnx_artifacts.py @@ -0,0 +1,113 @@ +"""Public ONNX artifact sanitation and reproducible, confined file digests.""" + +from __future__ import annotations + +import hashlib +import json +from pathlib import Path, PureWindowsPath + + +def sha256(path: Path) -> str: + digest = hashlib.sha256() + with Path(path).open("rb") as handle: + for chunk in iter(lambda: handle.read(4 * 1024 * 1024), b""): + digest.update(chunk) + return digest.hexdigest() + + +def source_snapshot_sha256(directory: Path) -> dict[str, str]: + """Hash native weights, tokenizer and reader configuration, not reports.""" + patterns = ( + "config.json", + "matryoshka_config.json", + "modules.json", + "sentence_bert_config.json", + "config_sentence_transformers.json", + "*.safetensors", + "*.safetensors.index.json", + "pytorch_model*.bin", + "*.bin.index.json", + "classification_heads.pt", + "tokenizer*", + "vocab*", + "merges.txt", + "special_tokens_map.json", + "added_tokens.json", + "1_Pooling/config.json", + ) + files = { + path + for pattern in patterns + for path in directory.glob(pattern) + if path.is_file() + } + return {str(path.relative_to(directory)): sha256(path) for path in sorted(files)} + + +def prepare_export_directory(directory: Path, identity: dict) -> None: + """Bind all precisions/exits to one source before changing output files.""" + manifest = directory / "source_identity.json" + if manifest.exists(): + if json.loads(manifest.read_text()) != identity: + raise ValueError( + "Export source or exit inventory changed; use a separate directory" + ) + return + if directory.exists() and any(directory.iterdir()): + raise ValueError( + "Existing export has no source identity; use a separate directory" + ) + directory.mkdir(parents=True, exist_ok=True) + manifest.write_text(json.dumps(identity, indent=2) + "\n") + + +def strip_debug_annotations(message) -> None: + """Remove recursive exporter metadata, preserving external tensor locations.""" + if hasattr(message, "doc_string"): + message.doc_string = "" + if hasattr(message, "metadata_props"): + del message.metadata_props[:] + for field, value in message.ListFields(): + if field.message_type is None: + continue + for child in value if field.is_repeated else (value,): + strip_debug_annotations(child) + + +def external_data_sha256(message, graph_path: Path) -> dict[str, str]: + """Hash every referenced tensor file, including subgraphs and attributes. + + Absolute paths, traversal, Windows drive paths and symlink escapes are + rejected before any external file is read. Load the model with + ``load_external_data=False`` to retain the locations for this check. + """ + locations = set() + + def visit(item): + if hasattr(item, "external_data"): + entries = {entry.key: entry.value for entry in item.external_data} + if "location" in entries: + locations.add(entries["location"]) + for field, value in item.ListFields(): + if field.message_type is not None: + for child in value if field.is_repeated else (value,): + visit(child) + + visit(message) + root = Path(graph_path).parent.resolve() + result = {} + for name in sorted(locations): + location = Path(name) + windows = PureWindowsPath(name) + resolved = (root / location).resolve() + if ( + not name + or location.is_absolute() + or windows.drive + or ".." in location.parts + or ".." in windows.parts + or not resolved.is_relative_to(root) + ): + raise ValueError("External data must stay inside its artifact directory") + result[name] = sha256(resolved) + return result diff --git a/reproducibility/onnx/qualify_onnx_fp32.py b/reproducibility/onnx/qualify_onnx_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..b8dc4cefc69e972461ad0ada71b6495217cdbfa5 --- /dev/null +++ b/reproducibility/onnx/qualify_onnx_fp32.py @@ -0,0 +1,152 @@ +"""Private fixed-weight FP32 ONNX qualification; no model-quality selection.""" + +import argparse +import hashlib +import json +import time +from collections import Counter +from pathlib import Path + +import numpy as np +import onnxruntime as ort + + +def sha256(path): + digest = hashlib.sha256() + with Path(path).open("rb") as stream: + for block in iter(lambda: stream.read(8 << 20), b""): + digest.update(block) + return digest.hexdigest() + + +def probabilities(logits): + values = np.exp(logits - logits.max(-1, keepdims=True)) + return values / values.sum(-1, keepdims=True) + + +p = argparse.ArgumentParser() +p.add_argument("--model", type=Path, required=True) +p.add_argument("--reference", type=Path, required=True) +p.add_argument("--graph", type=Path, required=True) +p.add_argument("--output", type=Path, required=True) +p.add_argument("--lengths", type=int, nargs="*") +p.add_argument("--diagnostic", action="store_true") +p.add_argument("--batch-size", type=int, choices=[1]) +p.add_argument("--sequence", action="store_true") +a = p.parse_args() +weight = sha256(a.model / "model.safetensors") +refs = np.load(a.reference, allow_pickle=False) +reference_weight = (json.loads(str(refs["metadata"]))['weights_sha256'] + if a.sequence else str(refs["weights_sha256"])) +assert reference_weight == weight +record = { + "qualified": False, + "weights_sha256": weight, + "reference_sha256": sha256(a.reference), + "graph_sha256": sha256(a.graph), + "precision": "FP32 weights and computation, exact query-block attention", + "probability_atol": 2e-4, + "probability_rtol": 1e-4, + "exact_BIO_required": not a.sequence, + "exact_argmax_required": True, + "task": "sequence-classification" if a.sequence else "token-classification", + "runtime": ort.__version__, + "rows": [], + "diagnostic": a.diagnostic, + "static_batch_size": a.batch_size, + "scope": "numerical qualification, not independent quality or isolated latency benchmark", +} + + +def save(): + temporary = a.output.with_suffix(".tmp") + temporary.write_text(json.dumps(record, indent=2) + "\n") + temporary.replace(a.output) + + +options = ort.SessionOptions() +options.intra_op_num_threads = 8 +options.inter_op_num_threads = 1 +options.enable_profiling = True +options.profile_file_prefix = str(a.output.with_suffix("")) +session = None +failure = None +numerical_failures = [] +try: + session = ort.InferenceSession( + str(a.graph), options, + providers=[("ROCMExecutionProvider", { + "device_id": "0", "arena_extend_strategy": "kSameAsRequested", + })], + ) + assert session.get_providers()[0] == "ROCMExecutionProvider" + record["providers"] = session.get_providers() + if a.batch_size: + assert all(value.shape[0] == a.batch_size for value in session.get_inputs()) + try: + session.run(None, {"input_ids": refs['b2s2_ids'], "attention_mask": refs['b2s2_mask']}) + except ort.capi.onnxruntime_pybind11_state.InvalidArgument as error: + assert 'Expected: 1' in str(error), str(error) + else: + raise AssertionError('Unqualified batch was accepted') + record['unqualified_batch_rejected'] = True + for key in refs["cases"].tolist(): + ids, mask = refs[key + "_ids"], refs[key + "_mask"] + if a.batch_size and ids.shape[0] != a.batch_size: + continue + if a.lengths and ids.shape[1] not in a.lengths: + continue + start = time.perf_counter() + actual = session.run(None, {"input_ids": ids, "attention_mask": mask})[0] + elapsed = (time.perf_counter() - start) * 1000 + expected = refs[key + "_logits"] + assert actual.dtype == np.float32, actual.dtype + assert actual.shape == expected.shape + assert np.isfinite(actual).all(), key + valid = slice(None) if a.sequence else mask.astype(bool) + scores = probabilities(actual[valid]) + reference = probabilities(expected[valid]) + row = { + "case": key, + "valid_tokens": mask.sum(1).tolist(), + "max_logit_error": float(abs(actual[valid] - expected[valid]).max()), + "max_probability_error": float(abs(scores - reference).max()), + "argmax_flips": int((scores.argmax(-1) != reference.argmax(-1)).sum()), + "first_inference_ms": elapsed, + "probability_gate": bool(np.allclose(scores, reference, atol=2e-4, rtol=1e-4)), + } + record["rows"].append(row) + save() + print(json.dumps(row), flush=True) + if not row["probability_gate"] or row["argmax_flips"]: + numerical_failures.append(key) + if not a.diagnostic: + raise AssertionError(row) + assert record["rows"], "No selected qualification cases" + if numerical_failures: + raise AssertionError("Numerical gate failures: " + ", ".join(numerical_failures)) +except Exception as error: + failure = error + record["failure"] = str(error) +finally: + if session is not None: + profile = Path(session.end_profiling()) + nodes = [item for item in json.loads(profile.read_text()) + if item.get("cat") == "Node" and item.get("args", {}).get("provider")] + record["profile_sha256"] = sha256(profile) + record["profile_providers"] = dict(Counter(item["args"]["provider"] for item in nodes)) + attention = [item for item in nodes if "__blocked_attention_" in item.get("name", "") + and item["args"].get("op_name") in ("MatMul", "FusedMatMul", "Softmax")] + record["attention_execution"] = dict(Counter( + item["args"]["provider"] + ":" + item["args"]["op_name"] for item in attention)) + record["attention_on_gpu"] = bool(attention) and all( + item["args"]["provider"] == "ROCMExecutionProvider" for item in attention) + if failure is None and not record["attention_on_gpu"]: + failure = AssertionError("No confirmed GPU attention, or attention fell back to CPU") + record["failure"] = str(failure) + record["qualified"] = failure is None + save() +if failure is not None: + raise failure +print(json.dumps({"qualified": True, "cases": len(record["rows"]), + "attention_execution": record["attention_execution"]}), flush=True) diff --git a/reproducibility/onnx/requirements.txt b/reproducibility/onnx/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..a7d7234dabd5beae6d3f24b4b0355659ffd09896 --- /dev/null +++ b/reproducibility/onnx/requirements.txt @@ -0,0 +1,9 @@ +# Install a Torch 2.10 build appropriate for your CPU/CUDA/ROCm hardware first. +# These are the actual FP32 exporter versions; this ORT wheel is CPU-only. +numpy==2.5.3 +onnx==1.18.0 +onnxscript==0.7.2 +onnxruntime==1.23.2 +transformers==4.57.6 +tokenizers==0.22.2 +safetensors==0.8.0 diff --git a/reproducibility/onnx/rewrite_blocked_attention.py b/reproducibility/onnx/rewrite_blocked_attention.py new file mode 100644 index 0000000000000000000000000000000000000000..baa067b0ec58baa78b2b6d181d28cae7f1643c8f --- /dev/null +++ b/reproducibility/onnx/rewrite_blocked_attention.py @@ -0,0 +1,657 @@ +"""Bound FP32 self-attention score tensors by looping over queries, never keys. + +This preserves the export's separate Q/K scale operations and mask fill values. +It is an optional graph variant; numerical and execution-provider qualification +of a real model remains separate from successful graph rewriting. +""" + +import argparse +import copy +import json +from pathlib import Path + +import numpy as np +import onnx +from onnx import TensorProto, helper, numpy_helper +from rewrite_graph import ( + attention_window, + build_maps, + find_attention_blocks, + inverted_padding_fill, + scalar_value, +) + +INT32_ELEMENTS = 2**31 - 1 +DEFAULT_SCORE_BYTES = 512 * 1024 * 1024 +FLOAT32_BYTES = 4 +ATTENTION_RANK = 4 +MASK_RANK = 2 +MIN_OPSET = 13 + + +def _array(graph, producers, name): + value = next((x for x in graph.initializer if x.name == name), None) + if value is None: + node = producers.get(name) + if node is not None and node.op_type == "Constant": + value = next((a.t for a in node.attribute if a.name == "value"), None) + return None if value is None else numpy_helper.to_array(value) + + +def _ancestors(producers, name): + result = {} + pending = [name] + while pending: + value = pending.pop() + node = producers.get(value) + if node is not None and node.output[0] not in result: + result[node.output[0]] = node + pending.extend(node.input) + return result + + +def _mask_contract(graph, producers, name): + window = attention_window(graph, producers, name) + ancestors = _ancestors(producers, name) + for node in ancestors.values(): + if node.op_type == "Cast" and any( + a.name == "to" + and a.i not in {TensorProto.FLOAT, TensorProto.BOOL, TensorProto.INT64} + for a in node.attribute + ): + raise ValueError( + "Narrow integer or mixed precision mask casts are unsupported" + ) + if node.op_type == "Range" and ( + scalar_value(graph, producers, node.input[0]) != 0 + or scalar_value(graph, producers, node.input[2]) != 1 + ): + raise ValueError( + "Only zero-based unit-step self-attention positions are supported" + ) + + def position_view(value): + node = producers.get(value) + if node is None: + raise ValueError("Unknown position view") + if node.op_type == "Range": + return ["position"] + if node.op_type == "Cast" and not any( + a.name == "to" and a.i == TensorProto.INT64 for a in node.attribute + ): + raise ValueError("Position differences must remain exact integers") + if node.op_type in {"Identity", "Cast"}: + return position_view(node.input[0]) + dims = position_view(node.input[0]) + if node.op_type == "Unsqueeze": + axes = _array(graph, producers, node.input[1]) + if axes is not None: + rank = len(dims) + len(axes) + if any(not -rank <= int(axis) < rank for axis in axes): + raise ValueError("Position unsqueeze axis is out of range") + axes = sorted(int(axis) % rank for axis in axes) + if len(set(axes)) == len(axes): + for axis in axes: + dims.insert(axis, 1) + return dims + elif node.op_type == "Transpose": + perm = next( + (list(a.ints) for a in node.attribute if a.name == "perm"), None + ) + if perm is not None and sorted(perm) == list(range(len(dims))): + return [dims[axis] for axis in perm] + elif node.op_type == "Reshape": + shape = _array(graph, producers, node.input[1]) + if ( + shape is not None + and list(shape).count(-1) == 1 + and all(x in {1, -1} for x in shape) + ): + return ["position" if x == -1 else 1 for x in shape] + raise ValueError("Unsupported positional broadcast") + + def predicate(value): + if value == "attention_mask": + return {"padding_2d"} + node = producers.get(value) + if node is None: + array = _array(graph, producers, value) + if array is not None and array.size == 1 and bool(array.item()): + return {"true"} + raise ValueError(f"Unknown attention-mask predicate {value!r}") + if node.op_type in {"Identity", "Expand", "Cast"}: + if node.op_type == "Cast" and not any( + a.name == "to" and a.i == TensorProto.BOOL for a in node.attribute + ): + raise ValueError("Positive mask predicate must remain boolean") + return predicate(node.input[0]) + if node.op_type == "Unsqueeze": + flags = predicate(node.input[0]) + axes = _array(graph, producers, node.input[1]) + if flags == {"padding_2d"}: + if axes is None or list(axes) != [1, 2]: + raise ValueError("Padding must broadcast over keys, not queries") + return {"padding"} + if "padding_2d" in flags: + raise ValueError("Unsupported padding broadcast") + return flags + if node.op_type == "And": + return predicate(node.input[0]) | predicate(node.input[1]) + if node.op_type in {"Less", "LessOrEqual"}: + # The shared recognizer proves abs(q-k) and the radius. Also prove + # these are the query and key axes, not two views of the same axis. + absolute = producers[node.input[0]] + difference = producers[absolute.input[0]] + views = [position_view(value) for value in difference.input] + views = [tuple([1] * (4 - len(view)) + view) for view in views] + if set(views) != {(1, 1, "position", 1), (1, 1, 1, "position")}: + raise ValueError( + "Local mask must compare absolute query and key positions" + ) + return {"window"} + if node.op_type == "GreaterOrEqual": + position = producers.get(node.input[0]) + while position is not None and position.op_type in { + "Cast", + "Identity", + "Unsqueeze", + "Squeeze", + "Reshape", + "Transpose", + }: + position = producers.get(position.input[0]) + if ( + scalar_value(graph, producers, node.input[1]) == 0 + and position is not None + and position.op_type == "Range" + and scalar_value(graph, producers, position.input[0]) == 0 + and scalar_value(graph, producers, position.input[2]) == 1 + ): + return {"true"} + raise ValueError("Unproved non-negative attention position predicate") + if node.op_type == "ConstantOfShape": + fill = next((a.t for a in node.attribute if a.name == "value"), None) + if fill is not None and numpy_helper.to_array(fill).item() == 1: + return {"true"} + if node.op_type == "Constant": + array = _array(graph, producers, value) + if array is not None and array.size == 1 and bool(array.item()): + return {"true"} + raise ValueError(f"Unknown attention-mask operation {node.op_type!r}") + + mask = producers[name] + first = scalar_value(graph, producers, mask.input[1]) + last = scalar_value(graph, producers, mask.input[2]) + if first == 0: + flags = predicate(mask.input[0]) + expected = {"padding"} | ({"window"} if window[0] >= 0 else set()) + if flags - {"true"} != expected: + raise ValueError( + "Attention mask is not the supported key-padding/window contract" + ) + return window[0], last, last, ancestors + if inverted_padding_fill(graph, producers, mask): + return -1, first, first, ancestors + condition = producers.get(mask.input[0]) + if condition is None or condition.op_type != "Not": + raise ValueError("Unknown negative-fill attention predicate") + inner_radius, padding_fill, _, _ = _mask_contract(graph, producers, mask.input[2]) + if inner_radius != -1 or predicate(condition.input[0]) != {"window"}: + raise ValueError("Unsupported nested local attention mask") + return window[0], padding_fill, first, ancestors + + +class _Nodes: + def __init__(self, prefix): + self.prefix = prefix + self.nodes = [] + + def op(self, kind, inputs, tag, **attributes): + name = self.prefix + tag + self.nodes.append( + helper.make_node(kind, inputs, [name], name=name, **attributes) + ) + return name + + def constant(self, tag, value, dtype=np.int64): + return self.op( + "Constant", + [], + tag, + value=numpy_helper.from_array(np.asarray(value, dtype=dtype)), + ) + + +def _replacement(block, index, radius, padding_fill, local_fill, block_size, elements): + outer = _Nodes(f"__blocked_attention_{index}_") + zero = outer.constant("zero", 0) + one = outer.constant("one", 1) + two = outer.constant("two", 2) + axes0 = outer.constant("axes0", [0]) + axes2 = outer.constant("axes2", [2]) + true = outer.constant("true", True, np.bool_) + float_zero = outer.constant("float_zero", 0, np.float32) + q = block["q_mul"].output[0] + k = block["k_mul"].output[0] + v = block["v_tensor"] + qshape = outer.op("Shape", [q], "qshape") + kshape = outer.op("Shape", [k], "kshape") + vshape = outer.op("Shape", [v], "vshape") + mshape = outer.op("Shape", ["attention_mask"], "mshape") + + def dimension(shape, axis, tag): + return outer.op( + "Gather", [shape, outer.constant(tag + "_axis", axis)], tag, axis=0 + ) + + batch = dimension(qshape, 0, "batch") + heads = dimension(qshape, 1, "heads") + queries = dimension(qshape, 2, "queries") + depth = dimension(qshape, 3, "depth") + keys = dimension(kshape, 3, "keys") + values = dimension(vshape, 3, "values") + bh = outer.op("Mul", [batch, heads], "batch_heads") + bhk = outer.op("Mul", [bh, keys], "batch_heads_keys") + denominator = outer.op("Max", [bhk, one], "nonzero_denominator") + capacity = outer.op( + "Div", [outer.constant("element_limit", elements), denominator], "capacity" + ) + valid = outer.op("Greater", [capacity, zero], "capacity_valid") + for tag, value in [ + ("batch", batch), + ("heads", heads), + ("queries", queries), + ("keys", keys), + ]: + positive = outer.op("Greater", [value, zero], tag + "_positive") + valid = outer.op("And", [valid, positive], tag + "_valid") + for tag, shape, axis, expected in [ + ("kb", kshape, 0, batch), + ("kh", kshape, 1, heads), + ("kd", kshape, 2, depth), + ("vb", vshape, 0, batch), + ("vh", vshape, 1, heads), + ("vk", vshape, 2, keys), + ("mb", mshape, 0, batch), + ("mk", mshape, 1, keys), + ]: + actual = dimension(shape, axis, tag) + equal = outer.op("Equal", [actual, expected], tag + "_equal") + valid = outer.op("And", [valid, equal], tag + "_valid") + equal_length = outer.op("Equal", [queries, keys], "self_attention_length") + valid = outer.op("And", [valid, equal_length], "shape_valid") + integer_mask = outer.op( + "Cast", ["attention_mask"], "integer_mask", to=TensorProto.INT64 + ) + is_zero = outer.op("Equal", [integer_mask, zero], "mask_zero") + is_one = outer.op("Equal", [integer_mask, one], "mask_one") + binary = outer.op("Or", [is_zero, is_one], "binary_mask") + binary = outer.op("Cast", [binary], "binary_integer", to=TensorProto.INT64) + binary = outer.op("ReduceMin", [binary], "all_binary", keepdims=0) + binary = outer.op("Equal", [binary, one], "binary_valid") + valid = outer.op("And", [valid, binary], "input_valid") + # ONNX has no Assert. An invalid shape deterministically fails Reshape before + # score allocation; never divide by zero or silently exceed the memory cap. + guard_size = outer.op("Where", [valid, one, two], "guard_size") + guard_shape = outer.op("Unsqueeze", [guard_size, axes0], "guard_shape") + guard = outer.op( + "Reshape", [outer.constant("guard_token", [1]), guard_shape], "guard" + ) + guard_scalar = outer.op("Gather", [guard, zero], "guard_scalar", axis=0) + requested = outer.constant("requested_block", block_size) + limited = outer.op("Min", [requested, capacity], "limited_block") + safe = outer.op("Max", [limited, one], "nonzero_block") + chunk = outer.op("Mul", [safe, guard_scalar], "block") + ceiling = outer.op("Add", [queries, chunk], "ceiling_sum") + ceiling = outer.op("Sub", [ceiling, one], "ceiling_numerator") + trips = outer.op("Div", [ceiling, chunk], "trips") + padding = outer.op("Cast", ["attention_mask"], "padding", to=TensorProto.BOOL) + padding = outer.op( + "Unsqueeze", [padding, outer.constant("pad_axes", [1, 2])], "padding4d" + ) + padding_bias = outer.op( + "Where", + [padding, float_zero, outer.constant("padding_fill", padding_fill, np.float32)], + "padding_bias", + ) + key_positions = ( + outer.op("Range", [zero, keys, one], "key_positions") if radius >= 0 else None + ) + + body = _Nodes(outer.prefix + "body_") + iteration, condition = body.prefix + "iteration", body.prefix + "condition" + start = body.op("Mul", [iteration, chunk], "start") + end = body.op("Add", [start, chunk], "unclipped_end") + end = body.op("Min", [end, queries], "end") + start_vector = body.op("Unsqueeze", [start, axes0], "start_vector") + end_vector = body.op("Unsqueeze", [end, axes0], "end_vector") + sliced_q = body.op("Slice", [q, start_vector, end_vector, axes2], "queries") + scores = body.op("MatMul", [sliced_q, k], "scores") + bias = padding_bias + if radius >= 0: + positions = body.op("Range", [start, end, one], "query_positions") + positions = body.op( + "Unsqueeze", [positions, body.constant("column_axis", [1])], "query_column" + ) + keys_row = body.op("Unsqueeze", [key_positions, axes0], "key_row") + difference = body.op("Sub", [positions, keys_row], "difference") + distance = body.op("Abs", [difference], "distance") + keep = body.op( + "LessOrEqual", [distance, body.constant("radius", radius)], "local_keep" + ) + bias = body.op( + "Where", + [keep, padding_bias, body.constant("local_fill", local_fill, np.float32)], + "local_bias", + ) + masked = body.op("Add", [scores, bias], "masked") + probability = body.op("Softmax", [masked], "probability", axis=-1) + if block["probability_nodes"]: + isnan = body.op("IsNaN", [probability], "isnan") + probability = body.op( + "Where", [isnan, float_zero, probability], "guarded_probability" + ) + result = body.op("MatMul", [probability, v], "result") + length = body.op("Sub", [end, start], "length") + remaining = body.op("Sub", [chunk, length], "remaining") + remaining = body.op("Unsqueeze", [remaining, axes0], "remaining_vector") + pads = body.op( + "Concat", + [ + body.constant("six_zeros", [0] * 6), + remaining, + body.constant("last_zero", [0]), + ], + "pads", + axis=0, + ) + scan = body.op("Pad", [result, pads, float_zero], "scan") + continued = body.op("Identity", [condition], "continue") + body_graph = helper.make_graph( + body.nodes, + body.prefix, + [ + helper.make_tensor_value_info(iteration, TensorProto.INT64, []), + helper.make_tensor_value_info(condition, TensorProto.BOOL, []), + ], + [ + helper.make_tensor_value_info(continued, TensorProto.BOOL, []), + helper.make_tensor_value_info(scan, TensorProto.FLOAT, [None] * 4), + ], + ) + scanned = outer.op("Loop", [trips, true], "scan", body=body_graph) + ordered = outer.op("Transpose", [scanned], "ordered", perm=[1, 2, 0, 3, 4]) + dimensions = [ + outer.op("Unsqueeze", [value, axes0], tag + "_vector") + for tag, value in [("b", batch), ("h", heads), ("v", values)] + ] + shape = outer.op( + "Concat", + [ + dimensions[0], + dimensions[1], + outer.constant("infer_length", [-1]), + dimensions[2], + ], + "result_shape", + axis=0, + ) + flat = outer.op("Reshape", [ordered, shape], "flat") + length_vector = outer.op("Unsqueeze", [queries, axes0], "query_length_vector") + outer.nodes.append( + helper.make_node( + "Slice", + [flat, outer.constant("slice_start", [0]), length_vector, axes2], + [block["output_tensor"]], + name=outer.prefix + "output", + ) + ) + return outer.nodes + + +def _dependencies(node): + values = set(node.input) + for attribute in node.attribute: + if attribute.type != onnx.AttributeProto.GRAPH: + continue + graph = attribute.g + defined = {x.name for x in graph.input} | {x.name for x in graph.initializer} + defined.update(output for child in graph.node for output in child.output) + used = {name for child in graph.node for name in _dependencies(child)} + values.update(used - defined) + return values + + +def _prune(graph): + producers, _ = build_maps(graph) + needed = {x.name for x in graph.output} + pending = list(needed) + while pending: + node = producers.get(pending.pop()) + if node is not None: + for name in _dependencies(node) - needed: + needed.add(name) + pending.append(name) + nodes = [n for n in graph.node if any(x in needed for x in n.output)] + del graph.node[:] + graph.node.extend(nodes) + initializers = [x for x in graph.initializer if x.name in needed] + del graph.initializer[:] + graph.initializer.extend(initializers) + infos = [x for x in graph.value_info if x.name in needed] + del graph.value_info[:] + graph.value_info.extend(infos) + return needed + + +def rewrite_model(model, block_size=256, max_score_bytes=DEFAULT_SCORE_BYTES): + """Return a new, validated model and receipt; do not mutate input on failure.""" + if ( + block_size not in {256, 512} + or not FLOAT32_BYTES <= max_score_bytes <= INT32_ELEMENTS * FLOAT32_BYTES + ): + raise ValueError( + "Use block cap 256/512 and a positive FP32 score budget below int32" + ) + if next((o.version for o in model.opset_import if not o.domain), 0) < MIN_OPSET: + raise ValueError("Blocked attention requires ONNX opset >=13") + graph = model.graph + if any( + n.op_type + in { + "Loop", + "Attention", + "MultiHeadAttention", + "FlashAttention", + "CKFlashAttention", + } + or n.domain not in {"", "ai.onnx"} + or n.name.startswith("__blocked_attention_") + for n in graph.node + ): + raise ValueError("Refusing an already blocked or control-flow model") + producers, _ = build_maps(graph) + blocks = find_attention_blocks(graph, producers) + if not blocks: + raise ValueError("No fully recognized attention blocks") + info = { + x.name: x.type.tensor_type + for x in [*graph.input, *graph.value_info, *graph.output] + } + floating = {TensorProto.FLOAT16, TensorProto.BFLOAT16, TensorProto.DOUBLE} + constant_types = [ + a.t.data_type + for n in graph.node + for a in n.attribute + if a.type == onnx.AttributeProto.TENSOR + ] + if ( + any(x.data_type in floating for x in graph.initializer) + or any(dtype in floating for dtype in constant_types) + or any(t.elem_type in floating for t in info.values()) + ): + raise ValueError("Only FP32 weights and attention are supported") + if any( + n.op_type == "Cast" + and any(a.name == "to" and a.i in floating for a in n.attribute) + for n in graph.node + ): + raise ValueError("Mixed precision attention graphs are refused") + mask_info = info.get("attention_mask") + if ( + mask_info is None + or mask_info.elem_type not in {TensorProto.INT64, TensorProto.INT32} + or len(mask_info.shape.dim) != MASK_RANK + ): + raise ValueError("Expected a rank-two integer attention_mask") + matched = {block["softmax"].output[0] for block in blocks} + for node in graph.node: + if node.op_type == "Softmax" and node.output[0] not in matched: + tensor = info.get(node.output[0]) + if tensor is None or len(tensor.shape.dim) != MASK_RANK: + raise ValueError( + "Unrecognized attention Softmax; refusing a partial rewrite" + ) + plans, removed, quadratic = [], set(), set() + for block in blocks: + for name in [ + block["q_mul"].output[0], + block["k_mul"].output[0], + block["v_tensor"], + ]: + tensor = info.get(name) + if ( + tensor is None + or tensor.elem_type != TensorProto.FLOAT + or len(tensor.shape.dim) != ATTENTION_RANK + ): + raise ValueError( + f"Attention tensor {name!r} needs explicit rank-four FP32 value_info" + ) + if any( + a.name == "axis" and a.i not in {-1, 3} for a in block["softmax"].attribute + ): + raise ValueError("Softmax must normalize the complete key dimension") + radius, pad_fill, local_fill, ancestors = _mask_contract( + graph, producers, block["mask_tensor"] + ) + plans.append((radius, pad_fill, local_fill)) + old = [ + block[x] for x in ["qk_matmul", "add_mask", "softmax", "av_matmul"] + ] + block["probability_nodes"] + names = {x.output[0] for x in old} + for node in old: + if node is block["av_matmul"]: + continue + if any( + output in {x.name for x in graph.output} for output in node.output + ) or any( + n.output[0] not in names + and any(output in n.input for output in node.output) + for n in graph.node + ): + raise ValueError("Attention intermediates have additional consumers") + removed.update(names) + quadratic.update( + name + for name, node in ancestors.items() + if node.op_type + in {"Where", "And", "Less", "LessOrEqual", "Abs", "Sub", "Expand"} + ) + result = copy.deepcopy(model) + replacement = { + b["output_tensor"]: _replacement( + b, i, *plans[i], block_size, max_score_bytes // FLOAT32_BYTES + ) + for i, b in enumerate(blocks) + } + nodes = [] + for node in result.graph.node: + if node.output[0] in replacement: + nodes.extend(replacement[node.output[0]]) + elif node.output[0] not in removed: + nodes.append(node) + del result.graph.node[:] + result.graph.node.extend(nodes) + live = _prune(result.graph) + if quadratic & live: + raise ValueError("Original dense mask computation has external consumers") + onnx.checker.check_model(result) + receipt = { + "attention_blocks": len(blocks), + "query_block_cap": block_size, + "maximum_score_bytes": max_score_bytes, + "score_elements_below_int32": True, + "precision": "float32", + "keys_and_values": "complete input for each query block", + "q_k_scaling": "original separate operations preserved", + "windows": [p[0] for p in plans], + "real_model_qualification": False, + } + return result, receipt + + +def specialize_batch(model, batch_size): + """Restrict a token-ID export to an explicitly qualified batch size.""" + if ( + isinstance(batch_size, bool) + or not isinstance(batch_size, int) + or batch_size < 1 + ): + raise ValueError("Static batch size must be a positive integer") + if {value.name for value in model.graph.input} != {"input_ids", "attention_mask"}: + raise ValueError( + "Static batch specialization requires token IDs and padding only" + ) + for value in model.graph.input: + tensor = value.type.tensor_type + if len(tensor.shape.dim) != MASK_RANK: + raise ValueError("Token IDs and padding must have rank two") + batch = tensor.shape.dim[0] + if batch.HasField("dim_value") and batch.dim_value != batch_size: + raise ValueError("Cannot change a pre-existing fixed batch contract") + result = copy.deepcopy(model) + for value in result.graph.input: + value.type.tensor_type.shape.dim[0].dim_value = batch_size + # Keep inferred intermediates symbolic; the runtime propagates the fixed + # input shape. No operator, weight, or attention computation is changed. + onnx.checker.check_model(result) + return result + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("input", type=Path) + parser.add_argument("output", type=Path) + parser.add_argument("--block-size", type=int, choices=[256, 512], default=256) + parser.add_argument("--max-score-bytes", type=int, default=DEFAULT_SCORE_BYTES) + parser.add_argument( + "--batch-size", + type=int, + help="Restrict token-ID inputs to this separately qualified static batch size", + ) + args = parser.parse_args() + if args.output.exists() or Path(str(args.output) + ".data").exists(): + raise ValueError("Refusing to overwrite an existing graph or external weights") + model, receipt = rewrite_model( + onnx.load(args.input), args.block_size, args.max_score_bytes + ) + if args.batch_size is not None: + model = specialize_batch(model, args.batch_size) + receipt["static_batch_size"] = args.batch_size + args.output.parent.mkdir(parents=True, exist_ok=True) + onnx.save_model( + model, + args.output, + save_as_external_data=True, + all_tensors_to_one_file=True, + location=args.output.name + ".data", + size_threshold=1024, + ) + onnx.checker.check_model(str(args.output)) + Path(str(args.output) + ".json").write_text(json.dumps(receipt, indent=2) + "\n") + print(json.dumps(receipt)) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/onnx/rewrite_graph.py b/reproducibility/onnx/rewrite_graph.py new file mode 100644 index 0000000000000000000000000000000000000000..68784059cb27a7d43534e6a5aa0894d9cbd6c514 --- /dev/null +++ b/reproducibility/onnx/rewrite_graph.py @@ -0,0 +1,1028 @@ +#!/usr/bin/env python3 +""" +Rewrite an mmBERT ONNX graph to replace the dense attention subgraph +with a single com.ck::CKFlashAttention custom-op node per layer. + +Key optimisations over the naive rewrite: + - Sliding-window (local) attention is handled via the CK kernel's built-in + window_size_left / window_size_right parameters instead of materialising a + dense [1, 1, seq, seq] mask tensor. + - The 2-D attention mask subgraph (Expand, Where, etc.) is replaced by a + lightweight 1-D padding bias [B, 1, 1, seq] derived directly from the + attention_mask input. This reduces mask memory from O(n^2) to O(n). +""" + +import argparse +import math +import os +import sys + +import numpy as np +import onnx +from onnx import TensorProto, helper, numpy_helper +from stable_pooling import stabilize_mean_pooling + +WHERE_INPUT_COUNT = 3 +MASKED_ATTENTION_MAX = -10000.0 +ROPE_FREQUENCY_RANK = 3 # [1, rotary_dimension / 2, 1] + + +def build_maps(graph): + out2node = {} + name2node = {} + for n in graph.node: + name2node[n.name] = n + for o in n.output: + out2node[o] = n + return out2node, name2node + + +def scalar_value(graph, out2node, name): + """Read a scalar export constant without materialising any model weights.""" + tensor = next((t for t in graph.initializer if t.name == name), None) + if tensor is not None and math.prod(tensor.dims) == 1: + return float(numpy_helper.to_array(tensor).item()) + node = out2node.get(name) + if node is None: + return None + if node.op_type == "Constant": + for attr in node.attribute: + if attr.name == "value" and math.prod(attr.t.dims) == 1: + return float(numpy_helper.to_array(attr.t).item()) + if node.op_type == "Cast": + return scalar_value(graph, out2node, node.input[0]) + return None + + +def attention_probability_path(graph, out2node, softmax): + """Accept only direct probabilities or Where(IsNaN(p), scalar_zero, p).""" + probability = softmax.output[0] + extra = [] + for node in graph.node: + if node.op_type != "Where" or len(node.input) != WHERE_INPUT_COUNT: + continue + guard = out2node.get(node.input[0]) + if ( + node.input[2] == probability + and guard is not None + and guard.op_type == "IsNaN" + and list(guard.input) == [probability] + and scalar_value(graph, out2node, node.input[1]) == 0 + ): + probability = node.output[0] + extra = [guard, node] + break + consumers = [ + n for n in graph.node if n.op_type == "MatMul" and n.input[0] == probability + ] + return consumers, extra + + +def find_attention_blocks(graph, out2node): + blocks = [] + softmaxes = [n for n in graph.node if n.op_type == "Softmax"] + + for sm in softmaxes: + add_mask = out2node.get(sm.input[0]) + if not add_mask or add_mask.op_type != "Add": + continue + + qk_mm = out2node.get(add_mask.input[0]) + if not qk_mm or qk_mm.op_type != "MatMul": + continue + + q_mul = out2node.get(qk_mm.input[0]) + if not q_mul or q_mul.op_type != "Mul": + continue + + k_mul = out2node.get(qk_mm.input[1]) + if not k_mul or k_mul.op_type != "Mul": + continue + + # K path: either direct Transpose (classifier) or Reshape->Transpose->Reshape (embedding) + k_transpose = None + k_extra_nodes = [] + k_input0 = out2node.get(k_mul.input[0]) + if k_input0 and k_input0.op_type == "Transpose": + k_transpose = k_input0 + elif k_input0 and k_input0.op_type == "Reshape": + inner = out2node.get(k_input0.input[0]) + if inner and inner.op_type == "Transpose": + k_transpose = inner + k_extra_nodes.append(k_input0) + pre_reshape = out2node.get(inner.input[0]) + if pre_reshape and pre_reshape.op_type == "Reshape": + k_extra_nodes.append(pre_reshape) + + if not k_transpose: + continue + + av_consumers, probability_nodes = attention_probability_path( + graph, out2node, sm + ) + if len(av_consumers) != 1: + continue + av_mm = av_consumers[0] + + # For K tensor: use input of the outermost Reshape (before transpose chain) + # or Transpose input directly if no wrapping Reshapes + if k_extra_nodes: + pre_transpose_reshape = out2node.get(k_transpose.input[0]) + k_tensor = ( + pre_transpose_reshape.input[0] + if ( + pre_transpose_reshape and pre_transpose_reshape.op_type == "Reshape" + ) + else k_transpose.input[0] + ) + else: + k_tensor = k_transpose.input[0] + + blocks.append( + { + "softmax": sm, + "add_mask": add_mask, + "qk_matmul": qk_mm, + "q_mul": q_mul, + "k_mul": k_mul, + "k_transpose": k_transpose, + "k_extra_nodes": k_extra_nodes, + "av_matmul": av_mm, + "probability_nodes": probability_nodes, + "q_tensor": q_mul.input[0], + "k_tensor": k_tensor, + "v_tensor": av_mm.input[1], + "mask_tensor": add_mask.input[1], + "output_tensor": av_mm.output[0], + } + ) + + return blocks + + +def compute_scale(graph, out2node, q_mul_node, k_mul_node=None, hdim=64): + q_scale = scalar_value(graph, out2node, q_mul_node.input[1]) + k_scale = scalar_value(graph, out2node, (k_mul_node or q_mul_node).input[1]) + if q_scale is not None and k_scale is not None: + return q_scale * k_scale + try: + scale_name = q_mul_node.input[1] + sqrt_node = out2node[scale_name] + cast_node = out2node[sqrt_node.input[0]] + div_node = out2node[cast_node.input[0]] + sqrt_hdim = out2node[div_node.input[1]] + hdim_node = out2node[sqrt_hdim.input[0]] + if hdim_node.op_type == "Constant": + hdim_val = float(numpy_helper.to_array(hdim_node.attribute[0].t)) + return 1.0 / math.sqrt(hdim_val) + except (KeyError, IndexError): + pass + return 1.0 / math.sqrt(hdim) + + +def inverted_padding_fill(graph, out2node, mask): + """Recognise HF's Where(bool(1-padding), negative, 1-padding) export. + + The broadcast axes are part of the contract: attention_mask [B,S] becomes + [B,1,1,S], so only key padding is masked. Do not accept arbitrary inversions. + """ + condition = out2node.get(mask.input[0]) + if ( + condition is None + or condition.op_type != "Cast" + or list(condition.input) != [mask.input[2]] + or not any( + a.name == "to" and a.i == TensorProto.BOOL for a in condition.attribute + ) + ): + return False + sub = out2node.get(mask.input[2]) + if ( + sub is None + or sub.op_type != "Sub" + or scalar_value(graph, out2node, sub.input[0]) != 1 + ): + return False + padding = out2node.get(sub.input[1]) + if padding is not None and padding.op_type == "Cast": + padding = out2node.get(padding.input[0]) + if padding is None or padding.op_type != "Expand": + return False + padding = out2node.get(padding.input[0]) + if ( + padding is None + or padding.op_type != "Unsqueeze" + or padding.input[0] != "attention_mask" + ): + return False + axes = next((t for t in graph.initializer if t.name == padding.input[1]), None) + if axes is None: + node = out2node.get(padding.input[1]) + if node is not None and node.op_type == "Constant": + axes = next((a.t for a in node.attribute if a.name == "value"), None) + return axes is not None and list(numpy_helper.to_array(axes)) == [1, 2] + + +def attention_window(graph, out2node, mask_tensor_name): + """Read symmetric positional windows from mask semantics, never names/counts. + + ModernBERT exports abs(query_position - key_position) <= local_attention/2. + Both endpoints are inclusive: a radius of 64 means (64, 64), not (63, 64). + Unknown comparisons are refused instead of silently becoming global attention. + """ + mask = out2node.get(mask_tensor_name) + if mask is None or mask.op_type != "Where" or len(mask.input) != WHERE_INPUT_COUNT: + raise ValueError(f"unsupported additive attention mask {mask_tensor_name!r}") + keep = scalar_value(graph, out2node, mask.input[1]) + masked = scalar_value(graph, out2node, mask.input[2]) + positive_condition = ( + keep == 0 and masked is not None and masked <= MASKED_ATTENTION_MAX + ) + negative_condition = False + if keep is not None and keep <= MASKED_ATTENTION_MAX: + negative_condition = inverted_padding_fill(graph, out2node, mask) + condition = out2node.get(mask.input[0]) + if condition is not None and condition.op_type == "Not": + window = out2node.get(condition.input[0]) + while window is not None and window.op_type == "Unsqueeze": + window = out2node.get(window.input[0]) + if window is not None and window.op_type in {"Less", "LessOrEqual"}: + # HF applies the window after the padding mask. A recursive + # global-mask check rejects reversed or otherwise unknown fills. + negative_condition = attention_window( + graph, out2node, mask.input[2] + ) == (-1, -1) + if not positive_condition and not negative_condition: + raise ValueError( + f"mask {mask_tensor_name!r} must map allowed positions to zero" + ) + ancestors = {} + pending = [mask_tensor_name] + leaves = set() + while pending: + name = pending.pop() + node = out2node.get(name) + if node is None: + leaves.add(name) + elif node.name not in ancestors: + ancestors[node.name] = node + pending.extend(node.input) + if "attention_mask" not in leaves: + raise ValueError(f"mask {mask_tensor_name!r} does not depend on attention_mask") + + def position_range(name): + node = out2node.get(name) + while node is not None: + if node.op_type in { + "Cast", + "Identity", + "Unsqueeze", + "Squeeze", + "Reshape", + } or ( + node.op_type == "Transpose" + and any( + a.name == "perm" and list(a.ints) == [1, 0] for a in node.attribute + ) + ): + node = out2node.get(node.input[0]) + else: + break + return node.output[0] if node is not None and node.op_type == "Range" else None + + radii = set() + for node in ancestors.values(): + if node.op_type not in { + "Less", + "LessOrEqual", + "Greater", + "GreaterOrEqual", + "Equal", + }: + continue + lhs = out2node.get(node.input[0]) + bound = scalar_value(graph, out2node, node.input[1]) + if ( + node.op_type == "GreaterOrEqual" + and bound == 0 + and position_range(node.input[0]) is not None + ): + continue # exporter bounds-checks non-negative query positions + if node.op_type in {"Less", "LessOrEqual"} and lhs is not None: + sub = out2node.get(lhs.input[0]) if lhs.op_type == "Abs" else None + if sub is not None and sub.op_type == "Sub" and bound is not None: + q_pos, k_pos = map(position_range, sub.input) + radius = bound - (1 if node.op_type == "Less" else 0) + if ( + q_pos is not None + and q_pos == k_pos + and radius >= 0 + and radius.is_integer() + ): + radii.add(int(radius)) + continue + raise ValueError(f"unsupported attention-mask comparison {node.name!r}") + if len(radii) > 1: + raise ValueError(f"mask {mask_tensor_name!r} has conflicting local windows") + return (next(iter(radii)),) * 2 if radii else (-1, -1) + + +def find_mask_only_nodes(graph, out2node): + """Find nodes that are exclusively part of the 2-D mask computation.""" + mask_roots = set() + for n in graph.node: + if n.op_type == "Where" and ( + n.name.startswith("/model/Where") or n.name.startswith("node_masked_fill") + ): + mask_roots.add(n.name) + + ancestors = set() + visited = set() + + def trace_back(name): + if name in visited: + return + visited.add(name) + if name in out2node: + n = out2node[name] + ancestors.add(n.name) + for inp in n.input: + trace_back(inp) + + for n in graph.node: + if n.name in mask_roots: + for inp in n.input: + trace_back(inp) + ancestors.update(mask_roots) + + mask_only = set() + for name in ancestors: + node = next((n for n in graph.node if n.name == name), None) + if not node: + continue + all_consumers_mask = True + for out in node.output: + consumers = [n2 for n2 in graph.node if out in n2.input] + if not all(c.name in ancestors for c in consumers): + all_consumers_mask = False + break + if all_consumers_mask: + mask_only.add(name) + + return mask_only + + +def create_1d_padding_bias_nodes(graph): + """ + Create ONNX nodes that derive a 1-D padding bias from attention_mask. + + attention_mask [B, seq] (int64, 1=valid 0=padding) + -> Cast to float16 + -> Sub(1.0, ...) = 1.0 for padding, 0.0 for valid + -> Mul(-65504.0, ...) = -65504 for padding, 0.0 for valid + -> Reshape to [B, 1, 1, seq] + + Returns (nodes_list, output_tensor_name). + """ + nodes = [] + prefix = "/model/_fa_pad_bias" + + # Cast attention_mask to float16 + cast_out = f"{prefix}/Cast_output_0" + nodes.append( + helper.make_node( + "Cast", + inputs=["attention_mask"], + outputs=[cast_out], + name=f"{prefix}/Cast", + to=TensorProto.FLOAT16, + ) + ) + + # Sub(1.0, cast_out) -> inverted mask + one_const = f"{prefix}/one" + nodes.append( + helper.make_node( + "Constant", + inputs=[], + outputs=[one_const], + name=f"{prefix}/Constant_one", + value=numpy_helper.from_array( + np.array(1.0, dtype=np.float16), name=one_const + ), + ) + ) + + sub_out = f"{prefix}/Sub_output_0" + nodes.append( + helper.make_node( + "Sub", inputs=[one_const, cast_out], outputs=[sub_out], name=f"{prefix}/Sub" + ) + ) + + # Mul(-65504.0, sub_out) -> padding positions get -65504 + neg_inf_const = f"{prefix}/neg_inf" + nodes.append( + helper.make_node( + "Constant", + inputs=[], + outputs=[neg_inf_const], + name=f"{prefix}/Constant_neg_inf", + value=numpy_helper.from_array( + np.array(-65504.0, dtype=np.float16), name=neg_inf_const + ), + ) + ) + + mul_out = f"{prefix}/Mul_output_0" + nodes.append( + helper.make_node( + "Mul", + inputs=[sub_out, neg_inf_const], + outputs=[mul_out], + name=f"{prefix}/Mul", + ) + ) + + # Reshape to [B, 1, 1, seq] -- use a constant shape [-1 is inferred] + # Actually use Unsqueeze with axes [1, 2] which is cleaner + axes_const = f"{prefix}/axes" + nodes.append( + helper.make_node( + "Constant", + inputs=[], + outputs=[axes_const], + name=f"{prefix}/Constant_axes", + value=numpy_helper.from_array( + np.array([1, 2], dtype=np.int64), name=axes_const + ), + ) + ) + + unsqueeze_out = f"{prefix}/Unsqueeze_output_0" + nodes.append( + helper.make_node( + "Unsqueeze", + inputs=[mul_out, axes_const], + outputs=[unsqueeze_out], + name=f"{prefix}/Unsqueeze", + ) + ) + + return nodes, unsqueeze_out + + +def rotary_fp32_initializers(graph, out2node): + """Identify FP32 RoPE frequencies used only with positions, then cast to FP16. + + Torch exports this intentional FP32 numerical island as an initializer. + It is not an encoder/head weight, and must not be narrowed to satisfy the + weight-name guard. Require the complete observed position-to-sin/cos path. + """ + consumers = {} + for node in graph.node: + for name in set(node.input): + consumers.setdefault(name, []).append(node) + graph_outputs = {v.name for v in graph.output} + + def only_consumer(name, op): + uses = consumers.get(name, []) + if name not in graph_outputs and len(uses) == 1 and uses[0].op_type == op: + return uses[0] + return None + + def attr(node, name): + return next( + (helper.get_attribute_value(a) for a in node.attribute if a.name == name), + None, + ) + + exempt = set() + for tensor in graph.initializer: + if ( + tensor.data_type != TensorProto.FLOAT + or len(tensor.dims) != ROPE_FREQUENCY_RANK + ): + continue + if tensor.dims[0] != 1 or tensor.dims[2] != 1: + continue + multiply = only_consumer(tensor.name, "MatMul") + if multiply is None or multiply.input[0] != tensor.name: + continue + position = out2node.get(multiply.input[1]) + if position is None or position.op_type != "Cast" or attr(position, "to") != 1: + continue + position = out2node.get(position.input[0]) + while position is not None and position.op_type == "Unsqueeze": + position = out2node.get(position.input[0]) + if ( + position is None + or position.op_type != "Range" + or scalar_value(graph, out2node, position.input[0]) != 0 + or scalar_value(graph, out2node, position.input[2]) != 1 + ): + continue + transpose = only_consumer(multiply.output[0], "Transpose") + if transpose is None or attr(transpose, "perm") != [0, 2, 1]: + continue + concat = only_consumer(transpose.output[0], "Concat") + if ( + concat is None + or list(concat.input) != [transpose.output[0]] * 2 + or attr(concat, "axis") not in (-1, 2) + or concat.output[0] in graph_outputs + ): + continue + trig = consumers.get(concat.output[0], []) + expected_trig = {"Cos", "Sin"} + if ( + len(trig) != len(expected_trig) + or {n.op_type for n in trig} != expected_trig + ): + continue + casts = [only_consumer(n.output[0], "Cast") for n in trig] + if all(n is not None and attr(n, "to") == TensorProto.FLOAT16 for n in casts): + exempt.add(tensor.name) + return exempt + + +def fp32_head_initializers(graph, out2node, blocks): + """Find FP32 constants confined to the post-attention output computation. + + No exempt tensor may contribute to any layer's Q, K or V. Unused weights + and graph outputs unrelated to attention are not a task head. This check + depends on dataflow, not initializer names or optional value_info. + """ + + def ancestors(names): + seen, pending = set(), list(names) + while pending: + name = pending.pop() + if name in seen: + continue + seen.add(name) + node = out2node.get(name) + if node is not None: + pending.extend(node.input) + return seen + + attention_outputs = {block["output_tensor"] for block in blocks} + attention_inputs = ancestors(attention_outputs) + head_paths = set() + for output in graph.output: + path = ancestors([output.name]) + if path & attention_outputs: + head_paths.update(path) + return { + tensor.name + for tensor in graph.initializer + if tensor.data_type == TensorProto.FLOAT + and tensor.name in head_paths + and tensor.name not in attention_inputs + } + + +def weight_precision( + initializers, *, position_constants=frozenset(), head_constants=frozenset() +): + """Return the one floating-point elem_type every weight tensor shares. + + Integer initializers (shapes, gather indices) carry no precision and are + ignored. A graph with no floating-point weights, with more than one + floating-point precision, or with a precision the CK kernel does not take + (only FLOAT and FLOAT16 are) is refused, since neither name would describe + it. + """ + counts = {} + for tensor in initializers: + if tensor.data_type in _FLOAT_TYPES and tensor.name not in ( + position_constants | head_constants + ): + counts[tensor.data_type] = counts.get(tensor.data_type, 0) + 1 + if not counts: + raise ValueError("the graph holds no floating-point weight tensors") + if len(counts) > 1: + found = ", ".join( + f"{n} {TensorProto.DataType.Name(t)}" for t, n in sorted(counts.items()) + ) + raise ValueError(f"the graph mixes weight precisions: {found}") + (elem_type,) = counts + if elem_type not in (TensorProto.FLOAT, TensorProto.FLOAT16): + raise ValueError( + f"the graph holds {TensorProto.DataType.Name(elem_type)} weights; " + "CKFlashAttention takes FLOAT or FLOAT16" + ) + return elem_type + + +_FLOAT_TYPES = ( + TensorProto.FLOAT, + TensorProto.FLOAT16, + TensorProto.BFLOAT16, + TensorProto.DOUBLE, +) + + +def output_precision_conflict(output_path, model_is_fp16): + """Return a message when the output name claims a precision the graph lacks. + + find_onnx_models ranks candidates by filename and puts model_fa_fp16.onnx + first whenever CK Flash Attention is available, so the name is a contract + rather than a label. The rewriter keeps the weights as it finds them and, + for an FP32 graph, only casts around each CKFlashAttention node. Running it + on model.onnx under an fp16 name is what produced the FP32 + model_fa_fp16.onnx in five of the six published 32K heads (#3256): a valid + graph with twice the memory of the file its name promises. + """ + name = os.path.basename(output_path).lower() + if "fp16" in name and not model_is_fp16: + return ( + f"{output_path}: the input graph holds FP32 weights, so this file would " + "hold FP32 weights under an fp16 name. Name an FP32 rewrite " + "model_fa.onnx; an fp16 name needs an FP16 input graph." + ) + return None + + +def enforce_output_precision(output_path, model_is_fp16): + """Exit on an fp16 name over FP32 weights; warn on the reverse.""" + conflict = output_precision_conflict(output_path, model_is_fp16) + if conflict: + print(f"error: {conflict}", file=sys.stderr) + sys.exit(1) + if model_is_fp16 and "fp16" not in os.path.basename(output_path).lower(): + print( + f"warning: {output_path} will hold FP16 weights but its name does not " + "say so; find_onnx_models ranks FP16 candidates by name." + ) + + +def rewrite( + model_path, output_path, hdim=64, local_attention=128, fp32_task_head=False +): + model = onnx.load(model_path) + graph = model.graph + + out2node, _name2node = build_maps(graph) + blocks = find_attention_blocks(graph, out2node) + + if not blocks: + print("No attention blocks found -- nothing to rewrite.") + sys.exit(1) + + softmax_count = sum(n.op_type == "Softmax" for n in graph.node) + if len(blocks) != softmax_count: + raise ValueError( + f"matched {len(blocks)} of {softmax_count} Softmax nodes; " + "refusing a partial rewrite that retains dense attention" + ) + windows = { + b["mask_tensor"]: attention_window(graph, out2node, b["mask_tensor"]) + for b in blocks + } + for window in windows.values(): + if window[0] >= 0 and window != (local_attention // 2,) * 2: + raise ValueError( + f"source mask window {window} disagrees with " + f"--local-attention={local_attention}" + ) + n_local = sum(windows[b["mask_tensor"]][0] >= 0 for b in blocks) + n_global = len(blocks) - n_local + print(f"Found {len(blocks)} attention blocks") + print( + f" Local attention layers: {n_local} (source windows={set(windows.values())})" + ) + print(f" Global attention layers: {n_global}") + + # Create 1-D padding bias nodes + pad_bias_nodes, pad_bias_tensor = create_1d_padding_bias_nodes(graph) + + # Detect model precision from the weights themselves. value_info is + # optional and describes activations, so a valid FP16 export that carries + # none would read as FP32, and one activation cannot vouch for every weight. + try: + position_constants = rotary_fp32_initializers(graph, out2node) + head_constants = ( + fp32_head_initializers(graph, out2node, blocks) if fp32_task_head else set() + ) + model_is_fp16 = ( + weight_precision( + graph.initializer, + position_constants=position_constants, + head_constants=head_constants, + ) + == TensorProto.FLOAT16 + ) + except ValueError as exc: + print(f"error: {exc}", file=sys.stderr) + sys.exit(1) + if model_is_fp16: + print(" Model precision: FP16 (skipping input/output Cast nodes)") + else: + print(" Model precision: FP32 (adding fp32↔fp16 Cast nodes)") + if position_constants: + print(f" Preserved {len(position_constants)} FP32 RoPE frequency constants") + if head_constants: + print(f" Preserved {len(head_constants)} FP32 post-attention head constants") + + enforce_output_precision(output_path, model_is_fp16) + + # Collect nodes to remove (attention subgraph) + nodes_to_remove = set() + new_nodes = [] + + for i, blk in enumerate(blocks): + scale = compute_scale(graph, out2node, blk["q_mul"], blk["k_mul"], hdim) + + for key in ( + "q_mul", + "k_mul", + "k_transpose", + "qk_matmul", + "add_mask", + "softmax", + "av_matmul", + ): + nodes_to_remove.add(blk[key].name) + for extra in blk.get("k_extra_nodes", []): + nodes_to_remove.add(extra.name) + for extra in blk["probability_nodes"]: + nodes_to_remove.add(extra.name) + + fa_wl, fa_wr = windows[blk["mask_tensor"]] + + # Derive a clean layer prefix for the FA node name + sm_name = blk["softmax"].name + if "/" in sm_name: + fa_name = sm_name.rsplit("/", 1)[0] + "/CKFlashAttention" + else: + fa_name = f"CKFlashAttention_{i}" + + if model_is_fp16: + # Model is already fp16: wire Q/K/V directly, output directly + fa_node = helper.make_node( + "CKFlashAttention", + inputs=[ + blk["q_tensor"], + blk["k_tensor"], + blk["v_tensor"], + pad_bias_tensor, + ], + outputs=[blk["output_tensor"]], + name=fa_name, + domain="com.ck", + scale=scale, + window_size_left=fa_wl, + window_size_right=fa_wr, + ) + new_nodes.append(fa_node) + else: + # Model is fp32: add Cast(fp32→fp16) for inputs, Cast(fp16→fp32) for output + q_fp16 = f"{fa_name}/q_cast_fp16" + k_fp16 = f"{fa_name}/k_cast_fp16" + v_fp16 = f"{fa_name}/v_cast_fp16" + out_fp16 = f"{fa_name}/out_fp16" + + new_nodes.append( + helper.make_node( + "Cast", + inputs=[blk["q_tensor"]], + outputs=[q_fp16], + name=f"{fa_name}/Cast_Q_fp16", + to=TensorProto.FLOAT16, + ) + ) + new_nodes.append( + helper.make_node( + "Cast", + inputs=[blk["k_tensor"]], + outputs=[k_fp16], + name=f"{fa_name}/Cast_K_fp16", + to=TensorProto.FLOAT16, + ) + ) + new_nodes.append( + helper.make_node( + "Cast", + inputs=[blk["v_tensor"]], + outputs=[v_fp16], + name=f"{fa_name}/Cast_V_fp16", + to=TensorProto.FLOAT16, + ) + ) + + fa_node = helper.make_node( + "CKFlashAttention", + inputs=[q_fp16, k_fp16, v_fp16, pad_bias_tensor], + outputs=[out_fp16], + name=fa_name, + domain="com.ck", + scale=scale, + window_size_left=fa_wl, + window_size_right=fa_wr, + ) + new_nodes.append(fa_node) + + new_nodes.append( + helper.make_node( + "Cast", + inputs=[out_fp16], + outputs=[blk["output_tensor"]], + name=f"{fa_name}/Cast_out_fp32", + to=TensorProto.FLOAT, + ) + ) + + # Remove dead scale-computation nodes + remaining_node_names = {n.name for n in graph.node} - nodes_to_remove + for blk in blocks: + scale_tensor = blk["q_mul"].input[1] + scale_node = out2node.get(scale_tensor) + if scale_node: + consumers = [ + n + for n in graph.node + if scale_tensor in n.input and n.name in remaining_node_names + ] + if not consumers: + nodes_to_remove.add(scale_node.name) + for inp in scale_node.input: + parent = out2node.get(inp) + if parent: + p_consumers = [ + n + for n in graph.node + if inp in n.input + and n.name in remaining_node_names + and n.name != scale_node.name + ] + if not p_consumers: + nodes_to_remove.add(parent.name) + remaining_node_names.discard(parent.name) + remaining_node_names.discard(scale_node.name) + + # DCE: remove 2-D mask subgraph nodes (Expand, Where, etc.) + # After CKFlashAttention nodes no longer reference Where_1/Where_2, + # the entire mask computation subgraph becomes dead code. + # Iteratively remove nodes whose outputs have no consumers. + all_new_node_names = {n.name for n in new_nodes} + all_pad_bias_names = {n.name for n in pad_bias_nodes} + kept_names = ( + remaining_node_names | all_new_node_names | all_pad_bias_names + ) - nodes_to_remove + + # Build a set of all tensor names consumed by new + pad-bias nodes + # so DCE doesn't remove their producers. + new_node_inputs = set() + for n in new_nodes + pad_bias_nodes: + new_node_inputs.update(n.input) + + graph_outputs = {o.name for o in graph.output} + changed = True + while changed: + changed = False + for n in list(graph.node): + if n.name not in kept_names or n.name in nodes_to_remove: + continue + if n.name in all_new_node_names or n.name in all_pad_bias_names: + continue + # Check if any output is consumed (by existing nodes OR new nodes) + all_dead = True + for out in n.output: + if out in graph_outputs: + all_dead = False + break + if out in new_node_inputs: + all_dead = False + break + consumers = [ + n2.name + for n2 in graph.node + if out in n2.input and n2.name in kept_names + ] + if consumers: + all_dead = False + break + if all_dead: + nodes_to_remove.add(n.name) + kept_names.discard(n.name) + changed = True + + # Rebuild node list: keep original order, then append new nodes + kept = [n for n in graph.node if n.name not in nodes_to_remove] + if any(output in windows for node in kept for output in node.output): + raise ValueError( + "dense attention mask is still live after rewriting every layer" + ) + kept.extend(pad_bias_nodes) + kept.extend(new_nodes) + + n_removed = len(graph.node) - len(kept) + len(pad_bias_nodes) + len(new_nodes) + orig_count = len(graph.node) + + del graph.node[:] + graph.node.extend(kept) + pooling_repairs = stabilize_mean_pooling(graph) + print(f" Stabilized {pooling_repairs} FP16 masked mean-pooling paths") + + # For fp16 models, cast graph outputs from fp16 to fp32 so that older + # ONNX Runtime host code (which only tries extract_tensor::) works. + if model_is_fp16: + output_cast_nodes = [] + for graph_out in graph.output: + if ( + graph_out.type.HasField("tensor_type") + and graph_out.type.tensor_type.elem_type == TensorProto.FLOAT16 + ): + old_name = graph_out.name + intermediate = f"{old_name}__fp16_raw" + # Rename the last node's output from old_name -> intermediate + for n in reversed(graph.node): + for idx, out in enumerate(n.output): + if out == old_name: + n.output[idx] = intermediate + break + else: + continue + break + # Also rename in value_info + for vi in graph.value_info: + if vi.name == old_name: + vi.name = intermediate + # Add Cast(fp16→fp32) as final output + output_cast_nodes.append( + helper.make_node( + "Cast", + inputs=[intermediate], + outputs=[old_name], + name=f"/model/_output_cast/{old_name}", + to=TensorProto.FLOAT, + ) + ) + # Update graph output type to fp32 + graph_out.type.tensor_type.elem_type = TensorProto.FLOAT + print(f" Added output Cast(fp16→fp32) for {old_name}") + graph.node.extend(output_cast_nodes) + + # Add com.ck opset import + has_ck = any(op.domain == "com.ck" for op in model.opset_import) + if not has_ck: + model.opset_import.append(helper.make_opsetid("com.ck", 1)) + + # Replacement nodes were appended after their original consumers. Keep the + # serialized graph valid without depending on an ORT-specific reordering. + available = ( + {v.name for v in graph.input} | {t.name for t in graph.initializer} | {""} + ) + pending = list(graph.node) + ordered = [] + while pending: + ready = [n for n in pending if all(name in available for name in n.input)] + if not ready: + raise ValueError("rewritten graph has unresolved inputs or a cycle") + for node in ready: + ordered.append(node) + available.update(node.output) + ready_names = {n.name for n in ready} + pending = [n for n in pending if n.name not in ready_names] + del graph.node[:] + graph.node.extend(ordered) + + onnx.save(model, output_path) + print(f"Saved rewritten model to {output_path}") + print(f" Removed {n_removed} nodes (attention + 2-D mask subgraph)") + print( + f" Added {len(new_nodes)} CKFlashAttention + {len(pad_bias_nodes)} padding-bias nodes" + ) + print(f" Total nodes: {len(graph.node)} (was {orig_count})") + + +def main(): + parser = argparse.ArgumentParser( + description="Rewrite mmBERT ONNX model to use CKFlashAttention custom op " + "with sliding-window masking and O(n) padding bias" + ) + parser.add_argument("input", help="Path to input ONNX model") + parser.add_argument("output", help="Path to output ONNX model") + parser.add_argument( + "--hdim", type=int, default=64, help="Head dimension (default: 64)" + ) + parser.add_argument( + "--local-attention", + type=int, + default=128, + help="Local attention window size (default: 128)", + ) + parser.add_argument( + "--fp32-task-head", + action="store_true", + help="Preserve FP32 head constants that cannot feed any attention Q/K/V", + ) + args = parser.parse_args() + rewrite( + args.input, args.output, args.hdim, args.local_attention, args.fp32_task_head + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/onnx/stable_pooling.py b/reproducibility/onnx/stable_pooling.py new file mode 100644 index 0000000000000000000000000000000000000000..03363b72caaf3552e2079200109bb545b1c3d6c0 --- /dev/null +++ b/reproducibility/onnx/stable_pooling.py @@ -0,0 +1,280 @@ +"""Keep an FP16 classifier's masked mean finite at long sequence lengths. + +Only the typed, mask-derived ReduceSum/Div path into a dense head is changed. +The encoder, mask multiplication, head, and public tensor dtypes stay intact. +Apply this after any whole-model FP16 conversion so its FP32 reduction and +division are not narrowed again. +""" + +from collections import defaultdict +from dataclasses import dataclass + +from onnx import NodeProto, TensorProto, helper, numpy_helper + + +def _standard(node, op): + return ( + node is not None + and node.op_type == op + and node.domain in ("", "ai.onnx") + and len(node.output) == 1 + ) + + +def _attribute(node, name, default=None): + return next( + (helper.get_attribute_value(a) for a in node.attribute if a.name == name), + default, + ) + + +class _GraphView: + def __init__(self, graph): + self.nodes = list(graph.node) + self.producers = {out: node for node in self.nodes for out in node.output} + self.consumers = defaultdict(list) + for node in self.nodes: + for name in node.input: + self.consumers[name].append(node) + self.initializers = {tensor.name: tensor for tensor in graph.initializer} + self.tensors = {} + for value in (*graph.input, *graph.value_info, *graph.output): + if value.type.HasField("tensor_type"): + tensor = value.type.tensor_type + if tensor.HasField("shape"): + shape = [ + ( + dim.dim_value + if dim.HasField("dim_value") + else dim.dim_param or None + ) + for dim in tensor.shape.dim + ] + self.tensors[value.name] = (tensor.elem_type, shape) + for tensor in graph.initializer: + self.tensors[tensor.name] = (tensor.data_type, list(tensor.dims)) + self.outputs = {value.name for value in graph.output} + self.names = set(self.tensors) | set(self.producers) + self.names.update(node.name for node in self.nodes) + + def typed(self, name, dtype, rank): + info = self.tensors.get(name) + return info is not None and info[0] == dtype and len(info[1]) == rank + + def axes(self, node, rank): + axes = _attribute(node, "axes") + # ReduceSum/Unsqueeze have an optional second, axes input. + if axes is None and len(node.input) == 2: # noqa: PLR2004 + tensor = self.initializers.get(node.input[1]) + if tensor is None: + producer = self.producers.get(node.input[1]) + if _standard(producer, "Constant"): + tensor = _attribute(producer, "value") + if ( + tensor is None + or tensor.data_type != TensorProto.INT64 + or list(tensor.dims) != [1] + ): + return None + try: + axes = numpy_helper.to_array(tensor).tolist() + except (ValueError, OSError): + return None + if axes is None or len(axes) != 1 or not -rank <= axes[0] < rank: + return None + return tuple(axis % rank for axis in axes) + + def fresh(self, base): + name, suffix = base, 0 + while name in self.names: + suffix += 1 + name = f"{base}_{suffix}" + self.names.add(name) + return name + + +@dataclass +class _MeanPool: + divide: NodeProto + numerator: NodeProto + denominator_cast: NodeProto + count: NodeProto + + +def _masked_product(view, product): + """Return the original rank-two integer mask, accepting either Mul order.""" + if ( + not _standard(product, "Mul") + or len(product.input) != 2 # noqa: PLR2004 - binary op + ): + return None + if not view.typed(product.output[0], TensorProto.FLOAT16, 3): + return None + for hidden, mask in (product.input, list(reversed(product.input))): + cast = view.producers.get(mask) + if ( + not view.typed(hidden, TensorProto.FLOAT16, 3) + or not _standard(cast, "Cast") + or len(cast.input) != 1 + or _attribute(cast, "to") != TensorProto.FLOAT16 + ): + continue + unsqueeze = view.producers.get(cast.input[0]) + if ( + not _standard(unsqueeze, "Unsqueeze") + or len(unsqueeze.input) not in (1, 2) + or view.axes(unsqueeze, 3) != (2,) + ): + continue + original_mask = unsqueeze.input[0] + if not view.typed(original_mask, TensorProto.INT64, 2): + continue + hidden_length = view.tensors[hidden][1][1] + mask_length = view.tensors[original_mask][1][1] + # A single-token mask broadcasting across the sequence is not this + # mean: its denominator would count one token, rather than the sequence. + if ( + hidden_length is not None + and mask_length is not None + and hidden_length != mask_length + ): + continue + return original_mask + return None + + +def _match_mean_pool(view, divide): + if ( + not _standard(divide, "Div") + or len(divide.input) != 2 # noqa: PLR2004 - binary op + or not view.typed(divide.output[0], TensorProto.FLOAT16, 2) + ): + return None + # This repair belongs to a sequence classifier, not an arbitrary attention + # normalization or a reduction feeding the encoder. + heads = view.consumers[divide.output[0]] + if not any( + _standard(head, "Gemm") + and len(head.input) >= 2 # noqa: PLR2004 - Gemm inputs A/B, optional C + and head.input[0] == divide.output[0] + and _attribute(head, "transA", 0) == 0 + and view.typed(head.input[1], TensorProto.FLOAT16, 2) + for head in heads + ): + return None + numerator = view.producers.get(divide.input[0]) + denominator_cast = view.producers.get(divide.input[1]) + if ( + not _standard(numerator, "ReduceSum") + or len(numerator.input) not in (1, 2) + or view.axes(numerator, 3) != (1,) + or _attribute(numerator, "keepdims", 1) != 0 + or not view.typed(numerator.output[0], TensorProto.FLOAT16, 2) + or not _standard(denominator_cast, "Cast") + or len(denominator_cast.input) != 1 + or _attribute(denominator_cast, "to") != TensorProto.FLOAT16 + or not view.typed(denominator_cast.output[0], TensorProto.FLOAT16, 2) + ): + return None + mask = _masked_product(view, view.producers.get(numerator.input[0])) + count = view.producers.get(denominator_cast.input[0]) + if ( + mask is None + or not _standard(count, "ReduceSum") + or len(count.input) not in (1, 2) + or count.input[0] != mask + or view.axes(count, 2) != (1,) + or _attribute(count, "keepdims", 1) != 1 + or not view.typed(count.output[0], TensorProto.INT64, 2) + ): + return None + return _MeanPool(divide, numerator, denominator_cast, count) + + +def stabilize_mean_pooling(graph): + """Repair recognized FP16 masked means in place; return the number changed. + + Required type/rank evidence comes from graph inputs, initializers and + value_info. Unrecognized, untyped and FP32 paths are left byte-for-byte + unchanged. Shared FP16 intermediates remain available to other consumers. + The original mean output name/type is preserved, making the pass idempotent. + """ + view = _GraphView(graph) + pools = [pool for node in view.nodes if (pool := _match_mean_pool(view, node))] + if not pools: + return 0 + + replacements, removed, new_info = {}, set(), [] + for pool in pools: + original_mean = pool.divide.output[0] + prefix = original_mean + "__stable_pool" + product = view.fresh(prefix + "_product_fp32") + total = view.fresh(prefix + "_sum_fp32") + count = view.fresh(prefix + "_count_fp32") + mean = view.fresh(prefix + "_mean_fp32") + nodes = [ + helper.make_node( + "Cast", + [pool.numerator.input[0]], + [product], + name=view.fresh(prefix + "_cast_product"), + to=TensorProto.FLOAT, + ), + ] + wider_sum = type(pool.numerator)() + wider_sum.CopyFrom(pool.numerator) + wider_sum.name = view.fresh(prefix + "_sum") + wider_sum.input[0] = product + wider_sum.output[0] = total + nodes.extend( + [ + wider_sum, + helper.make_node( + "Cast", + [pool.count.output[0]], + [count], + name=view.fresh(prefix + "_cast_count"), + to=TensorProto.FLOAT, + ), + helper.make_node( + "Div", [total, count], [mean], name=view.fresh(prefix + "_divide") + ), + helper.make_node( + "Cast", + [mean], + [original_mean], + name=view.fresh(prefix + "_cast_mean"), + to=TensorProto.FLOAT16, + ), + ] + ) + replacements[original_mean] = nodes + for node in (pool.numerator, pool.denominator_cast): + output = node.output[0] + if output not in view.outputs and view.consumers[output] == [pool.divide]: + removed.add(output) + for name, source in ( + (product, pool.numerator.input[0]), + (total, pool.numerator.output[0]), + (count, pool.denominator_cast.output[0]), + (mean, original_mean), + ): + new_info.append( + helper.make_tensor_value_info( + name, TensorProto.FLOAT, view.tensors[source][1] + ) + ) + + rewritten = [] + for node in view.nodes: + output = node.output[0] if len(node.output) == 1 else None + if output in replacements: + rewritten.extend(replacements[output]) + elif output not in removed: + rewritten.append(node) + kept_info = [value for value in graph.value_info if value.name not in removed] + del graph.node[:] + graph.node.extend(rewritten) + del graph.value_info[:] + graph.value_info.extend(kept_info + new_info) + return len(pools) diff --git a/reproducibility/onnx/test_rewrite_blocked_attention.py b/reproducibility/onnx/test_rewrite_blocked_attention.py new file mode 100644 index 0000000000000000000000000000000000000000..ba8bd3e8c9d222a051fbc79a8239e36451157fa1 --- /dev/null +++ b/reproducibility/onnx/test_rewrite_blocked_attention.py @@ -0,0 +1,431 @@ +"""Execute tiny FP32 ONNX attention graphs with dynamic batch/sequence shapes.""" + +import unittest + +import numpy as np +import onnx +from onnx import TensorProto, helper, numpy_helper +from rewrite_blocked_attention import rewrite_model, specialize_batch + +try: + import onnxruntime as ort +except ImportError: + ort = None + + +def attention_model(radius=None, guard=True, inverted=False, negative=-np.inf): + nodes, initializers = [], [] + + def const(name, value, dtype=np.int64): + initializers.append( + numpy_helper.from_array(np.asarray(value, dtype=dtype), name) + ) + return name + + def op(kind, inputs, name, **attrs): + nodes.append(helper.make_node(kind, inputs, [name], name=name, **attrs)) + return name + + zero = const("zero", 0, np.float32) + neg = const("negative", negative, np.float32) + const("q_scale", 0.37, np.float32) + const("k_scale", 0.29, np.float32) + const("pad_axes", [1, 2]) + const("zero_i", 0) + const("one_i", 1) + const("two_i", 2) + const("axes0", [0]) + const("axes1", [1]) + shape = op("Shape", ["q"], "q_shape") + length = op("Gather", [shape, "two_i"], "length", axis=0) + batch = op("Gather", [shape, "zero_i"], "batch", axis=0) + b = op("Unsqueeze", [batch, "axes0"], "batch_vector") + n = op("Unsqueeze", [length, "axes0"], "length_vector") + dense = op("Concat", [b, const("one_vector", [1]), n, n], "dense_shape", axis=0) + if inverted: + expanded = op("Unsqueeze", ["attention_mask", "pad_axes"], "expanded_padding") + expanded = op("Expand", [expanded, dense], "global_keep") + padding = op("Cast", [expanded], "float_padding", to=TensorProto.FLOAT) + inverse = op( + "Sub", [const("one_float", 1, np.float32), padding], "inverted_padding" + ) + is_pad = op("Cast", [inverse], "is_padding", to=TensorProto.BOOL) + mask = op("Where", [is_pad, neg, inverse], "global_mask") + else: + padding = op("Cast", ["attention_mask"], "bool_padding", to=TensorProto.BOOL) + expanded = op("Unsqueeze", [padding, "pad_axes"], "expanded_padding") + keep = op("Expand", [expanded, dense], "global_keep") + mask = op("Where", [keep, zero, neg], "global_mask") + if radius is not None: + position = op("Range", ["zero_i", length, "one_i"], "positions") + queries = op("Unsqueeze", [position, "axes1"], "query_positions") + keys = op("Unsqueeze", [position, "axes0"], "key_positions") + distance = op("Sub", [queries, keys], "difference") + distance = op("Abs", [distance], "distance") + keep = op("LessOrEqual", [distance, const("radius", radius)], "local_keep") + if inverted: + outside = op("Not", [keep], "outside") + # Different finite fills exercise exact nested mask reconstruction. + local_negative = const("local_negative", -100000000.0, np.float32) + mask = op("Where", [outside, local_negative, mask], "local_mask") + else: + keep = op("And", ["global_keep", keep], "combined_keep") + mask = op("Where", [keep, zero, neg], "local_mask") + kt = op("Transpose", ["k"], "kt", perm=[0, 1, 3, 2]) + qs = op("Mul", ["q", "q_scale"], "qs") + ks = op("Mul", [kt, "k_scale"], "ks") + scores = op("MatMul", [qs, ks], "scores") + scores = op("Add", [scores, mask], "masked") + probability = op("Softmax", [scores], "probability", axis=-1) + if guard: + isnan = op("IsNaN", [probability], "isnan") + probability = op("Where", [isnan, zero, probability], "guarded") + op("MatMul", [probability, "v"], "output") + inputs = [ + helper.make_tensor_value_info( + name, TensorProto.FLOAT, ["batch", 2, "sequence", 4] + ) + for name in ["q", "k", "v"] + ] + inputs.append( + helper.make_tensor_value_info( + "attention_mask", TensorProto.INT64, ["batch", "sequence"] + ) + ) + graph = helper.make_graph( + nodes, + "attention", + inputs, + [ + helper.make_tensor_value_info( + "output", TensorProto.FLOAT, ["batch", 2, "sequence", 4] + ) + ], + initializers, + ) + model = helper.make_model( + graph, opset_imports=[helper.make_opsetid("", 16)], ir_version=9 + ) + return onnx.shape_inference.infer_shapes(model) + + +def session(model): + options = ort.SessionOptions() + options.graph_optimization_level = ort.GraphOptimizationLevel.ORT_DISABLE_ALL + options.intra_op_num_threads = 1 + options.inter_op_num_threads = 1 + options.log_severity_level = 3 + return ort.InferenceSession( + model.SerializeToString(), options, providers=["CPUExecutionProvider"] + ) + + +def feeds(batch, length, all_masked=False): + rng = np.random.default_rng(770 + batch + length) + values = { + name: rng.normal(size=(batch, 2, length, 4)).astype(np.float32) + for name in ["q", "k", "v"] + } + mask = np.ones((batch, length), np.int64) + padding_stride = 3 + if length > padding_stride: + mask[0, 1::padding_stride] = ( + 0 # irregular padding exercises key-axis broadcasting + ) + if batch > 1: + mask[1, length // 2 :] = 0 + if all_masked: + mask[-1] = 0 + values["attention_mask"] = mask + return values + + +@unittest.skipIf(ort is None, "onnxruntime CPU is needed for execution tests") +class BlockedNumerics(unittest.TestCase): + def test_static_batch_rejects_unqualified_batches(self): + inputs = [ + helper.make_tensor_value_info( + name, TensorProto.INT64, ["batch", "sequence"] + ) + for name in ["input_ids", "attention_mask"] + ] + output = helper.make_tensor_value_info( + "output", TensorProto.INT64, ["batch", "sequence"] + ) + graph = helper.make_graph( + [helper.make_node("Mul", [value.name for value in inputs], ["output"])], + "batch_contract", + inputs, + [output], + ) + model = helper.make_model( + graph, opset_imports=[helper.make_opsetid("", 16)], ir_version=9 + ) + original = model.SerializeToString() + restricted = specialize_batch(model, 1) + self.assertEqual(model.SerializeToString(), original) + self.assertTrue( + all( + value.type.tensor_type.shape.dim[0].dim_value == 1 + for value in restricted.graph.input + ) + ) + runtime = session(restricted) + valid = {value.name: np.ones((1, 5), np.int64) for value in inputs} + np.testing.assert_array_equal( + runtime.run(None, valid)[0], np.ones((1, 5), np.int64) + ) + for wrong_input in ["input_ids", "attention_mask"]: + invalid = dict(valid) + invalid[wrong_input] = np.ones((2, 5), np.int64) + with self.assertRaises(Exception) as error: + runtime.run(None, invalid) + self.assertIn("Expected: 1", str(error.exception)) + with self.assertRaises(ValueError): + specialize_batch(restricted, 2) + with self.assertRaises(ValueError): + specialize_batch(model, 0) + with self.assertRaises(ValueError): + specialize_batch(attention_model(), 1) + + def test_dynamic_batch_tail_and_local_absolute_positions(self): + for radius, inverted in [(None, False), (2, False), (2, True)]: + with self.subTest(radius=radius, inverted=inverted): + original = attention_model(radius, inverted=inverted) + frozen = original.SerializeToString() + blocked, receipt = rewrite_model(original) + self.assertEqual(original.SerializeToString(), frozen) + self.assertEqual(receipt["attention_blocks"], 1) + reference, candidate = session(original), session(blocked) + for batch, length in [(1, 1), (2, 255), (1, 256), (2, 257), (1, 513)]: + inputs = feeds(batch, length) + before = reference.run(None, inputs)[0] + after = candidate.run(None, inputs)[0] + np.testing.assert_allclose(after, before, atol=2e-6, rtol=2e-6) + self.assertEqual(after.shape, (batch, 2, length, 4)) + + def test_adaptive_memory_limit_still_uses_complete_keys(self): + original = attention_model(radius=3) + blocked, _ = rewrite_model(original, block_size=512, max_score_bytes=16384) + inputs = feeds(2, 257) + # B*H*K=1028, hence the effective query block is three rows, not 512. + np.testing.assert_allclose( + session(blocked).run(None, inputs)[0], + session(original).run(None, inputs)[0], + atol=2e-6, + rtol=2e-6, + ) + + def test_all_masked_rows_preserve_nan_guard_and_finite_fill(self): + for negative, guard in [ + (-np.inf, True), + (-np.finfo(np.float32).max, True), + (-1000000000.0, False), + ]: + with self.subTest(negative=negative, guard=guard): + original = attention_model(guard=guard, negative=negative) + blocked, _ = rewrite_model(original) + inputs = feeds(2, 257, all_masked=True) + before = session(original).run(None, inputs)[0] + after = session(blocked).run(None, inputs)[0] + np.testing.assert_allclose(after, before, atol=2e-6, rtol=2e-6) + self.assertTrue(np.isfinite(after).all()) + if negative == -np.inf: + self.assertTrue((after[-1] == 0).all()) + + def test_nonbinary_padding_is_rejected(self): + blocked, _ = rewrite_model(attention_model()) + for invalid in [-1, 2]: + inputs = feeds(1, 17) + inputs["attention_mask"][0, 3] = invalid + with self.assertRaises(Exception) as error: + session(blocked).run(None, inputs) + self.assertIn("Reshape", str(error.exception)) + + def test_impossible_budget_fails_before_score_allocation(self): + blocked, _ = rewrite_model(attention_model(), max_score_bytes=16) + with self.assertRaises(Exception) as error: + session(blocked).run(None, feeds(2, 17)) + self.assertIn("Reshape", str(error.exception)) + + def test_mismatched_self_attention_shapes_are_rejected_at_runtime(self): + blocked, _ = rewrite_model(attention_model()) + inputs = feeds(1, 17) + inputs["attention_mask"] = np.ones((1, 16), np.int64) + with self.assertRaises(Exception) as error: + session(blocked).run(None, inputs) + self.assertIn("Reshape", str(error.exception)) + + +class BlockedContracts(unittest.TestCase): + def test_scales_survive_and_dense_mask_is_removed(self): + original = attention_model(2) + blocked, _ = rewrite_model(original) + original_nodes = {n.name: n for n in original.graph.node} + nodes = {n.name: n for n in blocked.graph.node} + for name in ["qs", "ks"]: + self.assertEqual( + nodes[name].SerializeToString(), + original_nodes[name].SerializeToString(), + ) + self.assertTrue(any(n.op_type == "Loop" for n in blocked.graph.node)) + for name in [ + "global_keep", + "global_mask", + "difference", + "distance", + "local_keep", + "local_mask", + ]: + self.assertNotIn(name, nodes) + onnx.checker.check_model(blocked, full_check=True) + + def test_unknown_mask_and_precision_fail_without_mutating_input(self): + for variant in [ + "or", + "query_padding", + "nonzero_positions", + "half_weight", + "missing_type", + ]: + with self.subTest(variant=variant): + model = attention_model(2) + if variant == "or": + next( + n for n in model.graph.node if n.name == "combined_keep" + ).op_type = "Or" + elif variant == "query_padding": + next( + x for x in model.graph.initializer if x.name == "pad_axes" + ).CopyFrom( + numpy_helper.from_array(np.array([1, 3], np.int64), "pad_axes") + ) + elif variant == "nonzero_positions": + next(n for n in model.graph.node if n.op_type == "Range").input[ + 0 + ] = "one_i" + elif variant == "half_weight": + model.graph.initializer.append( + numpy_helper.from_array(np.ones(2, np.float16), "half") + ) + else: + items = [x for x in model.graph.value_info if x.name != "qs"] + del model.graph.value_info[:] + model.graph.value_info.extend(items) + before = model.SerializeToString() + with self.assertRaises(ValueError): + rewrite_model(model) + self.assertEqual(model.SerializeToString(), before) + + def test_external_dense_mask_or_probability_consumers_are_refused(self): + for output in ["global_mask", "probability"]: + model = attention_model() + model.graph.output.append( + helper.make_tensor_value_info(output, TensorProto.FLOAT, [None] * 4) + ) + with self.assertRaises(ValueError): + rewrite_model(model) + + def test_position_bounds_require_unmodified_nonnegative_range(self): + for variant in ["valid", "nonzero_bound", "multiply", "offset", "narrow_cast"]: + with self.subTest(variant=variant): + model = attention_model(2) + index = next( + i + for i, n in enumerate(model.graph.node) + if n.name == "combined_keep" + ) + nodes = [] + position = "query_positions" + if variant in {"multiply", "offset"}: + nodes.append( + helper.make_node( + "Mul" if variant == "multiply" else "Add", + [position, "one_i"], + ["modified_positions"], + name="modified_positions", + ) + ) + position = "modified_positions" + elif variant == "narrow_cast": + nodes.append( + helper.make_node( + "Cast", + [position], + ["narrow_positions"], + name="narrow_positions", + to=TensorProto.INT8, + ) + ) + position = "narrow_positions" + bound = "one_i" if variant == "nonzero_bound" else "zero_i" + nodes.append( + helper.make_node( + "GreaterOrEqual", [position, bound], ["bounds"], name="bounds" + ) + ) + nodes.append( + helper.make_node( + "And", + ["local_keep", "bounds"], + ["bounded_keep"], + name="bounded_keep", + ) + ) + old = list(model.graph.node) + old[index].input[1] = "bounded_keep" + del model.graph.node[:] + model.graph.node.extend(old[:index] + nodes + old[index:]) + if variant == "valid": + rewrite_model(model) + else: + with self.assertRaises(ValueError): + rewrite_model(model) + + def test_local_position_axes_cannot_be_identical(self): + model = attention_model(2) + next(n for n in model.graph.node if n.name == "difference").input[ + 1 + ] = "query_positions" + with self.assertRaises(ValueError): + rewrite_model(model) + + def test_unmatched_or_partially_rewritten_attention_is_refused(self): + for kind, domain in [ + ("Softmax", ""), + ("Attention", ""), + ("CKFlashAttention", "com.ck"), + ]: + with self.subTest(kind=kind): + model = attention_model() + model.graph.node.append( + helper.make_node( + kind, ["q"], ["unmatched"], name="unmatched", domain=domain + ) + ) + model.graph.value_info.append( + helper.make_tensor_value_info( + "unmatched", TensorProto.FLOAT, [None] * 4 + ) + ) + before = model.SerializeToString() + with self.assertRaises(ValueError): + rewrite_model(model) + self.assertEqual(model.SerializeToString(), before) + + def test_invalid_options_and_already_blocked_are_refused(self): + for options in [ + {"block_size": 0}, + {"block_size": 1024}, + {"max_score_bytes": 0}, + {"max_score_bytes": 2**33}, + ]: + with self.assertRaises(ValueError): + rewrite_model(attention_model(), **options) + blocked, _ = rewrite_model(attention_model()) + with self.assertRaises(ValueError): + rewrite_model(blocked) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/prepare_data.py b/reproducibility/prepare_data.py new file mode 100644 index 0000000000000000000000000000000000000000..4ee2f2d8379f5e3ddd0b99280bba57595650e05e --- /dev/null +++ b/reproducibility/prepare_data.py @@ -0,0 +1,29 @@ +"""Rebuild Modality text-output contracts without downloading images or scoring a model.""" +import argparse,hashlib,json,subprocess,sys +from pathlib import Path + +def main(): + p=argparse.ArgumentParser();p.add_argument('--sources',type=Path,required=True);p.add_argument('--tokenizer',required=True);p.add_argument('--output',type=Path,required=True);a=p.parse_args();a.output.mkdir(parents=True,exist_ok=False) + package=Path('src/training/model_classifier/sequence_repair/data') + def run(module,*args):subprocess.run([sys.executable,'-m','src.training.model_classifier.'+module,*map(str,args)],check=True) + meta=json.loads((package/'aya-reviewed-labels-v1.json').read_text())['source'] + annotation=a.output/'annotations/aya-reviewed-labels-v1.jsonl' + run('sequence_repair.hydrate_annotations','--sidecar',package/'aya-reviewed-labels-v1.json','--source',a.sources/meta['repository'].replace('/','--')/meta['filename'],'--output',annotation) + labels=['biology','business','chemistry','computer science','economics','engineering','health','history','law','math','other','philosophy','physics','psychology'] + contract={'label2id':{v:k for k,v in enumerate(labels)},'id2label':{str(k):v for k,v in enumerate(labels)},'problem_type':'single_label_classification'} + (a.output/'domain-contract.json').write_text(json.dumps(contract,indent=2)+'\n') + run('domain_classifier.prepare_data','--sources',a.sources,'--legacy-datasets',a.sources,'--annotations',annotation,'--contract',a.output/'domain-contract.json','--output',a.output/'domain-v1') + run('modality_routing_classifier.prepare_vela_data','--sources',a.sources,'--domain-corpus',a.output/'domain-v1','--annotations',annotation,'--output',a.output/'modality-v1') + run('modality_routing_classifier.extend_contracts','--corpus',a.output/'modality-v1','--output',a.output/'modality-v2') + run('sequence_repair.prepare_context','--corpus',a.output/'modality-v2','--tokenizer',a.tokenizer,'--backgrounds',package/'task-context-backgrounds.json','--train-per-label','12','--output',a.output/'modality-v2-context-v2') + subset=a.output/'modality-v2-context-4k-v2';subset.mkdir() + for split in ['train','dev']: + parent=a.output/'modality-v2-context-v2'/f'{split}.jsonl' + (subset/f'{split}.jsonl').write_text(''.join(x+'\n' for x in parent.read_text().split('\n') if x and json.loads(x)['length_bucket']=='4096')) + checks=[] + for name,want in json.loads(Path('data-file-hashes.json').read_text()).items(): + got=hashlib.sha256((a.output/name).read_bytes()).hexdigest();checks.append({'file':name,'expected':want,'actual':got,'byte_equal':want==got}) + if want!=got:raise ValueError('Data checksum differs: '+name) + (a.output/'data-reproduction.json').write_text(json.dumps({'data_only':True,'files':checks},indent=2)+'\n') + print(json.dumps({'event':'complete','files':len(checks)}),flush=True) +if __name__=='__main__':main() diff --git a/reproducibility/release-lineage.json b/reproducibility/release-lineage.json new file mode 100644 index 0000000000000000000000000000000000000000..7a8d31253bcc2305868df2fe603792487da84aec --- /dev/null +++ b/reproducibility/release-lineage.json @@ -0,0 +1,8 @@ +{ + "final_stage": "modality-v6-vela-base", + "selected_step": 200, + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "all_predeclared_development_retention_gates_passed": true, + "test_used": false +} diff --git a/reproducibility/requirements.txt b/reproducibility/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..911d337a8cf6489dbb74353cc616116dd01d48c0 --- /dev/null +++ b/reproducibility/requirements.txt @@ -0,0 +1,9 @@ +transformers==4.57.6 +peft==0.18.1 +datasets==4.1.1 +accelerate==1.10.1 +pyarrow==25.0.1 +scikit-learn==1.7.2 +numpy==2.5.3 +huggingface-hub==0.36.0 +tokenizers==0.22.2 diff --git a/reproducibility/src/training/model_classifier/domain_classifier/README.md b/reproducibility/src/training/model_classifier/domain_classifier/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c42e57eff7786af1e01449bc471c2055423718d0 --- /dev/null +++ b/reproducibility/src/training/model_classifier/domain_classifier/README.md @@ -0,0 +1,54 @@ +# Vela Domain classifier data + +Domain is the academic-subject signal historically called `intent`. It does not +predict arbitrary user intents or select a downstream model by itself. The Vela +application name is `Vela-1.0-Encoder-307M-Domain`; runtime label IDs remain: + +| ID | Domain | ID | Domain | +| --- | --- | --- | --- | +| 0 | biology | 7 | history | +| 1 | business | 8 | law | +| 2 | chemistry | 9 | math | +| 3 | computer science | 10 | other | +| 4 | economics | 11 | philosophy | +| 5 | engineering | 12 | physics | +| 6 | health | 13 | psychology | + +`prepare_data.py` maps Global-MMLU subjects to this existing ontology, keeps the +question and answer choices, and excludes the answer key. English source +questions and their translations stay together. If normalized text connects two +source groups, all their translated siblings are unioned transitively before +splitting. Conflicting-label components are excluded. + +All available MMLU-Pro test/validation rows and the category supplement are used +for historical exposure screening. Exact matches or English unigram/bigram TF-IDF +cosine similarity of at least 0.75 force the whole translated group into training. +The screen is deliberately conservative and does not establish absence from old +weights or base pretraining. It also does not turn a benchmark question into a +natural conversational request. Independently reviewed Aya requests are added +with contributor-group splits and reported as a separate source. + +Fetch the pinned files and hydrate Aya labels using the +[shared workflow](../sequence_repair/README.md), then prepare the corpus: + +```bash +python -m src.training.model_classifier.domain_classifier.prepare_data \ + --sources artifacts/vela/sources \ + --legacy-datasets artifacts/vela/sources \ + --annotations artifacts/vela/annotations/aya-reviewed-labels-v1.jsonl \ + --contract artifacts/vela/reference-domain/config.json \ + --output artifacts/vela/domain +``` + +The reference contract is the original 14-label classifier configuration. The +initial continuation experiment uses the original fixed Domain adapter on +`mmbert-32k-yarn@72a23a6640489471eb4ff7ad3ec5bc80af8a27de`; every candidate receipt +must identify its actual base and adapter hashes. A later Base migration requires +a development-only short/long comparison before freezing final lineage. + +Use source-balanced development selection to keep the small natural-request +source visible. Report per-class and per-language quality, not only the aggregate +academic score. Long context variants mark the requested task within independent +authored background narratives; they are controlled stress tests, not natural +long-document estimates. Independent test evaluation occurs only after the +candidate and base lineage are frozen. diff --git a/reproducibility/src/training/model_classifier/domain_classifier/__init__.py b/reproducibility/src/training/model_classifier/domain_classifier/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..01199295015c73174c3a879b18d96759f9a0e798 --- /dev/null +++ b/reproducibility/src/training/model_classifier/domain_classifier/__init__.py @@ -0,0 +1 @@ +"""Fixed-ontology domain classification recipes.""" diff --git a/reproducibility/src/training/model_classifier/domain_classifier/prepare_data.py b/reproducibility/src/training/model_classifier/domain_classifier/prepare_data.py new file mode 100644 index 0000000000000000000000000000000000000000..60c6beda127e9bc4500d6c2bac08459700ba6507 --- /dev/null +++ b/reproducibility/src/training/model_classifier/domain_classifier/prepare_data.py @@ -0,0 +1,255 @@ +"""Prepare multilingual domains with historical-source exposure screening.""" + +# ruff: noqa: PLC0415 +import argparse +import hashlib +import json +from collections import Counter +from pathlib import Path + +from ..sequence_repair.corpus import group_split, write_corpus +from ..sequence_repair.data import load_contract, normalized_text + +SUBJECT_GROUPS = { + "biology": [ + "anatomy", + "college_biology", + "high_school_biology", + "medical_genetics", + "virology", + ], + "business": [ + "business_ethics", + "management", + "marketing", + "professional_accounting", + "public_relations", + ], + "chemistry": ["college_chemistry", "high_school_chemistry"], + "computer science": [ + "college_computer_science", + "high_school_computer_science", + "computer_security", + "machine_learning", + ], + "economics": [ + "econometrics", + "high_school_macroeconomics", + "high_school_microeconomics", + ], + "engineering": ["electrical_engineering"], + "health": [ + "clinical_knowledge", + "college_medicine", + "professional_medicine", + "human_aging", + "human_sexuality", + "nutrition", + ], + "history": [ + "high_school_european_history", + "high_school_us_history", + "high_school_world_history", + "prehistory", + ], + "law": ["international_law", "jurisprudence", "professional_law"], + "math": [ + "abstract_algebra", + "college_mathematics", + "elementary_mathematics", + "high_school_mathematics", + "high_school_statistics", + ], + "philosophy": [ + "philosophy", + "moral_disputes", + "moral_scenarios", + "formal_logic", + "logical_fallacies", + ], + "physics": [ + "astronomy", + "college_physics", + "conceptual_physics", + "high_school_physics", + ], + "psychology": ["high_school_psychology", "professional_psychology"], +} +LANGUAGES = ["en", "zh", "es", "fr", "de", "ja"] +SIMILARITY_LIMIT = 0.75 +GLOBAL_REVISION = "0e619dbeb34206cd48705a1a0ea7fb21cae09993" + + +def question_with_choices(row): + return ( + row["question"] + + "\n" + + "\n".join(f"{letter.upper()}. {row[f'option_{letter}']}" for letter in "abcd") + ) + + +def canonical_groups(rows): + """Union translated siblings when any normalized request repeats.""" + parents = {row["group_id"]: row["group_id"] for row in rows} + + def root(group): + while parents[group] != group: + parents[group] = parents[parents[group]] + group = parents[group] + return group + + texts = {} + for row in rows: + key = normalized_text(row["text"]) + previous = texts.setdefault(key, row["group_id"]) + first, second = sorted([root(previous), root(row["group_id"])]) + parents[second] = first + labels = {} + for row in rows: + row["group_id"] = root(row["group_id"]) + labels.setdefault(row["group_id"], set()).add(row["label"]) + retained = [row for row in rows if len(labels[row["group_id"]]) == 1] + return retained, len(rows) - len(retained) + + +def main(): + from pyarrow import parquet + from sklearn.feature_extraction.text import TfidfVectorizer + from sklearn.neighbors import NearestNeighbors + + parser = argparse.ArgumentParser() + parser.add_argument("--sources", type=Path, required=True) + parser.add_argument("--legacy-datasets", type=Path, required=True) + parser.add_argument("--annotations", type=Path, required=True) + parser.add_argument("--contract", required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + labels, ids = load_contract(args.contract) + contract = { + "label2id": labels, + "id2label": ids, + "problem_type": "single_label_classification", + } + mapping = { + subject: label + for label, subjects in SUBJECT_GROUPS.items() + for subject in subjects + } + known = [] + for folder in [ + "TIGER-Lab--MMLU-Pro", + "LLM-Semantic-Router--category-classifier-supplement", + ]: + for path in (args.legacy_datasets / folder).rglob("*.parquet"): + for row in parquet.read_table(path).to_pylist(): + text = row.get("question", row.get("text")) + if text: + known.append(normalized_text(text)) + known = sorted(set(known)) + known_set = set(known) + if not known: + raise ValueError("Historical-source screening requires actual source rows") + english = parquet.read_table( + args.sources / "CohereLabs--Global-MMLU/en/test-00000-of-00001.parquet" + ).to_pylist() + vectorizer = TfidfVectorizer( + ngram_range=(1, 2), max_features=60000, min_df=2, sublinear_tf=True + ) + matrix = vectorizer.fit_transform(known) + search = NearestNeighbors( + n_neighbors=1, metric="cosine", algorithm="brute", n_jobs=8 + ).fit(matrix) + exposure = {} + for start in range(0, len(english), 128): + batch = english[start : start + 128] + distance, _ = search.kneighbors( + vectorizer.transform([normalized_text(row["question"]) for row in batch]) + ) + for index, row in enumerate(batch): + exposure[row["sample_id"]] = ( + normalized_text(row["question"]) in known_set + or 1 - float(distance[index, 0]) >= SIMILARITY_LIMIT + ) + groups = { + row["sample_id"]: "global-mmlu:" + + hashlib.sha256( + normalized_text(question_with_choices(row)).encode() + ).hexdigest() + for row in english + } + group_exposed = {} + for identifier, group in groups.items(): + group_exposed[group] = group_exposed.get(group, False) or exposure[identifier] + raw_rows = [] + for language in LANGUAGES: + rows = parquet.read_table( + args.sources + / f"CohereLabs--Global-MMLU/{language}/test-00000-of-00001.parquet" + ).to_pylist() + for row in rows: + raw_rows.append( + { + "id": f"global-{language}-{row['sample_id']}", + "text": question_with_choices(row), + "label": mapping.get(row["subject"], "other"), + "group_id": groups[row["sample_id"]], + "language": language, + "source": "global-mmlu-domain", + "subject": row["subject"], + "historical_source_similarity": bool(exposure[row["sample_id"]]), + "length_bucket": "short", + } + ) + raw_rows, quarantine = canonical_groups(raw_rows) + group_exposed = {} + for row in raw_rows: + group = row["group_id"] + group_exposed[group] = ( + group_exposed.get(group, False) or row["historical_source_similarity"] + ) + splits = {split: [] for split in ["train", "dev", "test"]} + counts = Counter() + for row in sorted( + raw_rows, key=lambda item: hashlib.sha256(item["id"].encode()).hexdigest() + ): + split = ( + "train" if group_exposed[row["group_id"]] else group_split(row["group_id"]) + ) + cap = {"train": 180, "dev": 25, "test": 35}[split] + key = (split, row["language"], row["label"]) + if counts[key] >= cap: + continue + counts[key] += 1 + splits[split].append({**row, "split": split}) + for line in args.annotations.read_text().split("\n"): + if not line.strip(): + continue + row = json.loads(line) + label = row["task_labels"]["domain"] + if label is None: + continue + split = group_split(row["group_id"]) + splits[split].append( + {**row, "label": label, "split": split, "length_bucket": "short"} + ) + provenance = { + "global_mmlu_revision": GLOBAL_REVISION, + "license": "Apache-2.0", + "historical_sources_screened": len(known), + "ambiguous_label_rows_quarantined": quarantine, + "screening": "Exact normalized text or word unigram/bigram TF-IDF cosine >=0.75 moves entire translated question group into train", + "historical_exposure_not_guaranteed": True, + "subject_mapping": SUBJECT_GROUPS, + "unmapped_subject_policy": "other", + "annotations_sha256": hashlib.sha256(args.annotations.read_bytes()).hexdigest(), + "source_group_rows": dict( + Counter(row["source"] for rows in splits.values() for row in rows) + ), + } + print( + json.dumps(write_corpus(args.output, splits, contract, provenance)), flush=True + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/modality_routing_classifier/extend_contracts.py b/reproducibility/src/training/model_classifier/modality_routing_classifier/extend_contracts.py new file mode 100644 index 0000000000000000000000000000000000000000..d2975d3941bdec8ec2285605afa173933f4fea36 --- /dev/null +++ b/reproducibility/src/training/model_classifier/modality_routing_classifier/extend_contracts.py @@ -0,0 +1,202 @@ +"""Train-only output-contract diversity; held-out wording and source groups stay frozen.""" + +# Language-specific punctuation is intentional. +# ruff: noqa: RUF001 +import argparse +import hashlib +import json +from pathlib import Path + +from ..sequence_repair.corpus import write_corpus +from ..sequence_repair.data import normalized_text +from .prepare_vela_data import TEMPLATES + +CONTRACTS = { + "en": { + "AR": [ + "An artist sent me their finished work. Draft accessible alternative text for it; a replacement artwork is unnecessary. Subject: {}", + "I am preparing a tutorial. Return a numbered set of written drawing instructions about this brief: {}", + "Return editable SVG source as plain text for the brief below. I will render the code myself. Brief: {}", + "Review the wording of this visual brief and suggest improvements in prose. The brief is quoted material: {}", + ], + "DIFFUSION": [ + "Deliver a finished raster artwork as an image file for this brief; drawing instructions would not satisfy the request: {}", + "Could you render the following scene? I only need the resulting visual asset, without a prose discussion: {}", + "Use your picture generator for the brief below. The deliverable is the picture itself: {}", + "A newly rendered image is the sole requested output. Here is the subject to depict: {}", + ], + "BOTH": [ + "I need an original rendered visual asset together with a separate written account of the design choices. Brief: {}", + "Please give me a picture file and an accompanying explanation outside the picture. Subject: {}", + "There are two deliverables for this brief: a generated artwork and a prose analysis of it. Brief: {}", + "Supply a fresh illustration plus accessible alt text as a separate written response. Subject: {}", + ], + }, + "zh": { + "AR": [ + "画师已经交付了成品，请为它撰写无障碍替代文本，不需要替换现有作品。主题：{}", + "我要编写教程，请把这个绘画需求整理成编号的文字步骤：{}", + "请返回可编辑的SVG源文本，我会自行渲染代码。设计要求：{}", + "请校对引号中的视觉设计说明，用文字提出措辞建议。这段说明只是引用材料：{}", + ], + "DIFFUSION": [ + "请交付完成的位图画作文件，文字绘画步骤不能满足这项要求。设计说明：{}", + "能否把这个场景渲染出来？我只需要最终视觉素材，不需要散文讨论：{}", + "请调用图片生成器处理下面的需求，交付物是图片本身：{}", + "唯一需要的输出是一张新渲染的图像，需要表现的主题是：{}", + ], + "BOTH": [ + "我需要原创渲染的视觉素材，同时另附一段解释设计选择的文字。设计要求：{}", + "请给出图片文件，并在图片之外提供配套说明。主题：{}", + "这个需求有两个交付物：生成的画作，以及对画作的文字分析。要求：{}", + "请提供全新的插画，再单独用文字给出无障碍替代描述。主题：{}", + ], + }, + "es": { + "AR": [ + "Un artista ya entregó su obra. Redacta un texto alternativo accesible; no hace falta sustituir la obra. Tema: {}", # codespell:ignore accesible + "Estoy preparando un tutorial. Devuelve una lista numerada de instrucciones escritas para dibujar: {}", + "Devuelve el código fuente SVG editable como texto. Yo mismo lo renderizaré. Encargo: {}", + "Revisa la redacción de este encargo visual citado y propón mejoras por escrito: {}", + ], + "DIFFUSION": [ + "Entrega una obra rasterizada terminada como archivo de imagen; las instrucciones para dibujar no bastan. Encargo: {}", + "Renderiza esta escena. Solo necesito el recurso visual resultante, sin comentario en prosa: {}", + "Utiliza el generador de imágenes para este encargo. El producto solicitado es la propia imagen: {}", # codespell:ignore utiliza + "El único resultado solicitado es una imagen recién renderizada que represente: {}", # codespell:ignore represente + ], + "BOTH": [ + "Necesito un recurso visual original renderizado junto con un texto separado sobre las decisiones de diseño: {}", + "Dame un archivo de imagen y una explicación adjunta fuera de la imagen. Tema: {}", + "Este encargo tiene dos entregables: una obra generada y un análisis escrito de ella: {}", + "Entrega una ilustración nueva y, en una respuesta escrita aparte, su texto alternativo accesible: {}", # codespell:ignore accesible + ], + }, + "fr": { + "AR": [ + "Un artiste a déjà livré son œuvre. Rédige un texte alternatif accessible sans remplacer cette œuvre. Sujet : {}", + "Je prépare un tutoriel. Fournis une liste numérotée d'instructions écrites pour dessiner : {}", # codespell:ignore tutoriel + "Renvoie le code source SVG modifiable sous forme de texte. Je le rendrai moi-même. Sujet : {}", + "Relis la formulation de ce brief visuel cité et propose des améliorations par écrit : {}", + ], + "DIFFUSION": [ + "Livre une œuvre matricielle terminée sous forme de fichier image ; des consignes de dessin ne suffisent pas. Sujet : {}", + "Peux-tu rendre cette scène ? Je souhaite uniquement le fichier visuel obtenu, sans commentaire en prose : {}", + "Utilise le générateur d'images pour ce brief. Le livrable demandé est l'image elle-même : {}", + "Le seul résultat demandé est une image nouvellement rendue représentant : {}", + ], + "BOTH": [ + "Il me faut un visuel original rendu ainsi qu'un texte séparé expliquant les choix graphiques : {}", + "Fournis un fichier image et une explication jointe, hors de l'image. Sujet : {}", + "Ce brief comporte deux livrables : une œuvre générée et son analyse écrite : {}", + "Livre une illustration nouvelle et son texte alternatif accessible dans une réponse écrite séparée : {}", + ], + }, + "de": { + "AR": [ + "Ein Künstler hat sein fertiges Werk bereits geliefert. Verfasse einen barrierefreien Alternativtext; ein Ersatzbild ist unnötig. Motiv: {}", + "Ich bereite eine Anleitung vor. Liefere nummerierte schriftliche Zeichenschritte zu diesem Motiv: {}", # codespell:ignore vor + "Gib bearbeitbaren SVG-Quelltext aus. Ich werde den Code selbst rendern. Vorgabe: {}", + "Prüfe die Formulierung dieser zitierten Bildvorgabe und schlage schriftliche Verbesserungen vor: {}", # codespell:ignore vor + ], + "DIFFUSION": [ + "Liefere ein fertiges Rasterbild als Bilddatei; schriftliche Zeichenanweisungen genügen nicht. Vorgabe: {}", # codespell:ignore als + "Rendere diese Szene. Ich brauche nur das fertige visuelle Ergebnis ohne begleitende Erörterung: {}", # codespell:ignore szene + "Verwende den Bildgenerator für diese Vorgabe. Das gewünschte Ergebnis ist das Bild selbst: {}", + "Als einzige Ausgabe wird ein neu gerendertes Bild dieses Motivs benötigt: {}", # codespell:ignore als + ], + "BOTH": [ + "Ich brauche ein neu gerendertes Motiv und zusätzlich einen getrennten Text über die Gestaltungsentscheidungen: {}", + "Liefere eine Bilddatei sowie eine begleitende Erklärung außerhalb des Bildes. Motiv: {}", + "Diese Vorgabe verlangt zwei Ergebnisse: ein erzeugtes Kunstwerk und eine schriftliche Analyse davon: {}", + "Liefere eine neue Illustration und ihren barrierefreien Alternativtext als getrennte schriftliche Antwort: {}", # codespell:ignore als + ], + }, + "ja": { + "AR": [ + "画家が完成した作品をすでに納品しました。作品を置き換えずに、アクセシビリティ用の代替テキストを書いてください。題材：{}", + "教材を準備しています。この題材の描き方を番号付きの文章の手順にしてください：{}", + "編集できるSVGソースをテキストで返してください。コードの描画はこちらで行います。要件：{}", + "引用された視覚デザインの説明文を校正し、表現の改善案を文章で提示してください：{}", + ], + "DIFFUSION": [ + "完成したラスター作品を画像ファイルで納品してください。描画手順の文章では要件を満たしません。要件：{}", + "この場面をレンダリングしてください。必要なのは完成した視覚素材だけで、文章の考察は不要です：{}", + "次の要件を画像生成器で処理してください。求める成果物は画像そのものです：{}", + "唯一必要な出力は、この題材を新たにレンダリングした画像です：{}", + ], + "BOTH": [ + "新しくレンダリングした視覚素材と、デザイン上の選択を説明する別の文章が必要です：{}", + "画像ファイルと、画像の外に添える説明文を提供してください。題材：{}", + "この依頼には二つの成果物があります。生成した作品と、その作品に関する文章の分析です：{}", + "新しいイラストを提供し、さらに別の文章の回答としてアクセシビリティ用の代替説明も書いてください：{}", + ], + }, +} + + +def recover_caption(row): + language = row["language"] + family = int(row["template_family"].rsplit("-", 1)[1]) + prefix = TEMPLATES[language]["AR"][family].split("{}")[0] + if not row["text"].startswith(prefix): + raise ValueError("Source caption wrapper differs from the frozen recipe") + caption = row["text"][len(prefix) :] + if ( + hashlib.sha256(normalized_text(caption).encode()).hexdigest() + != row["caption_sha256"] + ): + raise ValueError("Recovered caption differs from its frozen source hash") + return caption + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument("--corpus", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + splits = { + split: [ + json.loads(line) + for line in (args.corpus / f"{split}.jsonl").read_text().split("\n") + if line.strip() + ] + for split in ["train", "dev", "test"] + } + additions = [] + for row in splits["train"]: + if ( + row["source"] != "authored-output-contract-with-gallery-caption" + or row["label"] != "AR" + ): + continue + language = row["language"] + caption = recover_caption(row) + for label, templates in CONTRACTS[language].items(): + for index, template in enumerate(templates): + additions.append( + { + **row, + "id": f"contract-v2:{row['caption_sha256']}:{language}:{label}:{index}", + "text": template.format(caption), + "label": label, + "source": "authored-train-contract-diversity", + "template_family": f"modality-train-diversity-{index}", + } + ) + splits["train"].extend(additions) + contract = json.loads((args.corpus / "contract.json").read_text()) + provenance = { + "parent_manifest_sha256": hashlib.sha256( + (args.corpus / "manifest.json").read_bytes() + ).hexdigest(), + "training_additions": len(additions), + "scope": "Train-only authored output-contract variations on original train-only gallery source groups. Frozen dev/test requests and wording remain unchanged. Captions remain in their source language.", + "source_license": "CC0-1.0 for DiffusionDB captions", + "test_used": False, + } + print(json.dumps(write_corpus(args.output, splits, contract, provenance), indent=2)) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/modality_routing_classifier/prepare_vela_data.py b/reproducibility/src/training/model_classifier/modality_routing_classifier/prepare_vela_data.py new file mode 100644 index 0000000000000000000000000000000000000000..c6c288322bea245d20f31d8b50d0984a682c1d95 --- /dev/null +++ b/reproducibility/src/training/model_classifier/modality_routing_classifier/prepare_vela_data.py @@ -0,0 +1,298 @@ +"""Requested-output modality with task-matched text and image-prompt sources.""" + +# ruff: noqa: PLC0415, RUF001 +import argparse +import hashlib +import json +from collections import Counter +from pathlib import Path + +from ..sequence_repair.corpus import group_split, write_corpus +from ..sequence_repair.data import normalized_text + +MIN_CAPTION_CHARACTERS = 40 +MAX_CAPTION_CHARACTERS = 250 +HASH_SAMPLE_CUTOFF = 3 +LABELS = {"AR": 0, "DIFFUSION": 1, "BOTH": 2} +TEMPLATES = { + "en": { + "AR": [ + "Explain how I could draw this subject, using text only; do not generate an image: {}", + "Write Python code to draw this scene; return source code, not a rendered picture: {}", + "Translate this image prompt into French without generating an image: {}", + "I already have a picture of this subject. Suggest a text-only caption: {}", + "Discuss the visual composition described here. I want written analysis and no newly generated picture: {}", + "Turn this picture description into a paragraph of prose. Do not create an image: {}", + ], + "DIFFUSION": [ + "Generate an image of this scene: {}", + "Draw a picture based on the following description: {}", + "Create a digital illustration showing: {}", + "I need a new visual depicting the following; deliver the image alone: {}", + "Please produce a picture from this brief, with no written explanation: {}", + "Make a new piece of visual artwork showing this subject: {}", + ], + "BOTH": [ + "Explain the concept in writing and generate an image of: {}", + "Give me a written description and a newly generated illustration of: {}", + "Create a picture of this scene and add a short textual explanation: {}", + "Deliver two things: an explanatory paragraph and a new image depicting: {}", + "I want a written overview accompanied by a generated picture of this subject: {}", + "Produce both a visual illustration and separate explanatory text about: {}", + ], + }, + "zh": { + "AR": [ + "只用文字解释如何画这个主题，不要生成图片：{}", + "编写绘制这个场景的Python代码，只返回源代码，不要渲染图片：{}", + "把这个图片提示词翻译成英文，不要生成图片：{}", + "我已有这个主题的图片，请只提供文字图注：{}", + "分析以下视觉构图，只要书面分析，不需要新生成的图片：{}", + "把这段图片描述改写为散文，不要创建图像：{}", + ], + "DIFFUSION": [ + "生成描绘以下场景的图片：{}", + "根据以下描述画一张图：{}", + "创作一幅数字插画，内容是：{}", + "我需要一张全新的视觉作品，只交付图像：{}", + "请根据以下要求制作图片，不要附带文字解释：{}", + "为这个主题创作一幅新的视觉艺术作品：{}", + ], + "BOTH": [ + "用文字解释，并生成对应图片：{}", + "请提供一段书面描述和一张新生成的插画：{}", + "创作这个场景的图片，再给出简短文字说明：{}", + "请交付两部分：一个解释性段落和一张新图片，主题是：{}", + "我需要文字概述，以及配套生成的主题图片：{}", + "同时提供视觉插画和单独的文字解释：{}", + ], + }, + "es": { + "AR": [ + "Explica solo con texto cómo dibujar esto; no generes una imagen: {}", + "Escribe código Python para dibujar la escena; devuelve código, no una imagen renderizada: {}", + "Traduce esta descripción visual al inglés sin generar imágenes: {}", + "Ya tengo una imagen de este tema; sugiere solo un pie de foto escrito: {}", # codespell:ignore foto + "Analiza por escrito esta composición visual; no quiero una imagen nueva: {}", + "Convierte esta descripción de una imagen en un párrafo de prosa, sin crear imágenes: {}", + ], + "DIFFUSION": [ + "Genera una imagen de esta escena: {}", + "Dibuja una imagen basada en esta descripción: {}", + "Crea una ilustración digital que muestre: {}", + "Necesito una representación visual nueva; entrega solo la imagen: {}", + "Produce una imagen a partir de esta idea, sin explicación escrita: {}", + "Haz una nueva obra visual que represente este tema: {}", # codespell:ignore represente + ], + "BOTH": [ + "Explica el concepto por escrito y genera una imagen de: {}", + "Dame una descripción escrita y una ilustración nueva de: {}", + "Crea una imagen de esta escena y añade una explicación breve: {}", + "Entrega dos cosas: un párrafo explicativo y una imagen nueva de: {}", + "Quiero una visión general escrita acompañada de una imagen generada de: {}", + "Produce una ilustración visual y, por separado, un texto explicativo sobre: {}", + ], + }, + "fr": { + "AR": [ + "Explique uniquement par écrit comment dessiner ce sujet, sans générer d'image : {}", + "Écris du code Python pour dessiner cette scène ; renvoie le code et non une image rendue : {}", + "Traduis cette description visuelle en anglais sans produire d'image : {}", + "J'ai déjà une image de ce sujet ; propose seulement une légende écrite : {}", + "Analyse cette composition visuelle par écrit ; je ne veux pas de nouvelle image : {}", + "Transforme cette description d'image en un paragraphe de prose, sans créer d'image : {}", # codespell:ignore transforme + ], + "DIFFUSION": [ + "Génère une image de cette scène : {}", + "Dessine une image à partir de cette description : {}", + "Crée une illustration numérique représentant : {}", + "Il me faut une nouvelle représentation visuelle ; livre uniquement l'image : {}", + "Produis une image à partir de cette idée, sans explication écrite : {}", + "Réalise une nouvelle œuvre visuelle sur ce sujet : {}", + ], + "BOTH": [ + "Explique le concept par écrit et génère une image de : {}", + "Donne-moi une description écrite et une nouvelle illustration de : {}", + "Crée une image de cette scène et ajoute une courte explication : {}", + "Livre deux éléments : un paragraphe explicatif et une nouvelle image de : {}", + "Je veux une présentation écrite accompagnée d'une image générée de : {}", + "Produis à la fois une illustration et un texte explicatif séparé sur : {}", + ], + }, + "de": { + "AR": [ + "Erkläre nur in Worten, wie man dieses Motiv zeichnet; erzeuge kein Bild: {}", + "Schreibe Python-Code zum Zeichnen dieser Szene; liefere Quellcode statt eines gerenderten Bildes: {}", # codespell:ignore szene + "Übersetze diese Bildbeschreibung ins Englische, ohne ein Bild zu erzeugen: {}", + "Ich habe bereits ein Bild dieses Motivs; schlage nur eine schriftliche Bildunterschrift vor: {}", # codespell:ignore vor + "Analysiere diese Bildkomposition schriftlich; ich möchte kein neues Bild: {}", + "Formuliere diese Bildbeschreibung als Prosaabsatz, ohne ein Bild zu erstellen: {}", # codespell:ignore als + ], + "DIFFUSION": [ + "Erzeuge ein Bild dieser Szene: {}", # codespell:ignore szene + "Zeichne ein Bild nach dieser Beschreibung: {}", + "Erstelle eine digitale Illustration von: {}", + "Ich brauche eine neue visuelle Darstellung; liefere nur das Bild: {}", + "Erzeuge zu dieser Vorgabe ein Bild ohne schriftliche Erklärung: {}", + "Schaffe ein neues visuelles Kunstwerk zu diesem Motiv: {}", + ], + "BOTH": [ + "Erkläre das Konzept schriftlich und erzeuge ein Bild von: {}", + "Gib mir eine schriftliche Beschreibung und eine neue Illustration von: {}", + "Erstelle ein Bild dieser Szene und ergänze eine kurze Erklärung: {}", # codespell:ignore szene + "Liefere zwei Dinge: einen erklärenden Absatz und ein neues Bild von: {}", + "Ich möchte einen schriftlichen Überblick zusammen mit einem erzeugten Bild von: {}", + "Erstelle sowohl eine Illustration als auch einen separaten Erklärungstext zu: {}", # codespell:ignore als + ], + }, + "ja": { + "AR": [ + "この題材の描き方を文章だけで説明してください。画像は生成しないでください：{}", + "この場面を描くPythonコードを書き、画像ではなくソースコードだけを返してください：{}", + "この画像の説明文を英語に翻訳してください。画像は生成しないでください：{}", + "この題材の画像はすでに持っています。文章のキャプションだけを提案してください：{}", + "この視覚的構図を文章で分析してください。新しい画像は不要です：{}", + "この画像の説明を散文の段落に書き換えてください。画像は作成しないでください：{}", + ], + "DIFFUSION": [ + "次の場面の画像を生成してください：{}", + "次の説明に基づいて絵を描いてください：{}", + "次の内容を描いたデジタルイラストを作成してください：{}", + "新しい視覚表現が必要です。画像だけを納品してください：{}", + "この依頼内容から画像を作り、文章の説明は付けないでください：{}", + "この題材を表現する新しい視覚芸術作品を制作してください：{}", + ], + "BOTH": [ + "概念を文章で説明し、対応する画像も生成してください：{}", + "文章による説明と新しいイラストを両方提供してください：{}", + "この場面の画像を作成し、短い文章の説明も加えてください：{}", + "説明文の段落と新しい画像という二つの成果物を提供してください：{}", + "文章による概要と、それに添える生成画像が欲しいです：{}", + "視覚的なイラストと別の説明文を同時に制作してください：{}", + ], + }, +} + + +def main(): + from pyarrow import parquet + + parser = argparse.ArgumentParser() + parser.add_argument("--sources", type=Path, required=True) + parser.add_argument("--domain-corpus", type=Path, required=True) + parser.add_argument("--annotations", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + candidates = {} + source = parquet.ParquetFile( + args.sources / "poloclub--diffusiondb/metadata.parquet" + ) + for batch in source.iter_batches(batch_size=8192, columns=["prompt", "user_name"]): + for row in batch.to_pylist(): + prompt = row["prompt"].strip() + key = hashlib.sha256(normalized_text(prompt).encode()).hexdigest() + if ( + not MIN_CAPTION_CHARACTERS <= len(prompt) <= MAX_CAPTION_CHARACTERS + or int(key[:2], 16) > HASH_SAMPLE_CUTOFF + or not row["user_name"] + ): + continue + candidates.setdefault( + key, + { + "text": prompt, + "group_id": "diffusiondb-author-" + + hashlib.sha256(str(row["user_name"]).encode()).hexdigest(), + }, + ) + splits = {split: [] for split in ["train", "dev", "test"]} + counts = Counter() + for key, row in sorted(candidates.items()): + split = group_split(row["group_id"]) + if counts[split] >= {"train": 200, "dev": 40, "test": 60}[split]: + continue + index = counts[split] + counts[split] += 1 + family = {"train": [0, 1, 2], "dev": [3], "test": [4, 5]}[split] + family = family[index % len(family)] + for language, labels in TEMPLATES.items(): + for label, templates in labels.items(): + splits[split].append( + { + "id": f"modality-{key}-{language}-{label}", + "text": templates[family].format(row["text"]), + "label": label, + "group_id": row["group_id"], + "language": language, + "source": "authored-output-contract-with-gallery-caption", + "split": split, + "length_bucket": "short", + "caption_sha256": key, + "template_family": f"modality-{split}-{family}", + } + ) + for split in ["train", "dev", "test"]: + count = Counter() + for line in (args.domain_corpus / f"{split}.jsonl").read_text().split("\n"): + if not line.strip(): + continue + row = json.loads(line) + if ( + row["source"] != "global-mmlu-domain" + or count[row["language"]] + >= {"train": 120, "dev": 30, "test": 40}[split] + ): + continue + count[row["language"]] += 1 + splits[split].append( + { + **row, + "id": "modality-text-" + row["id"], + "label": "AR", + "source": "academic-text-answer", + } + ) + for line in args.annotations.read_text().split("\n"): + if not line.strip(): + continue + row = json.loads(line) + label = row["task_labels"]["modality"] + if label is None: + continue + split = group_split(row["group_id"]) + splits[split].append( + { + **row, + "id": "modality-" + row["id"], + "label": label, + "split": split, + "length_bucket": "short", + } + ) + contract = { + "label2id": LABELS, + "id2label": {index: label for label, index in LABELS.items()}, + "problem_type": "single_label_classification", + } + provenance = { + "diffusiondb_revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "gallery_license": "CC0-1.0", + "other_source_licenses": ["Apache-2.0"], + "domain_parent_manifest_sha256": hashlib.sha256( + (args.domain_corpus / "manifest.json").read_bytes() + ).hexdigest(), + "annotation_roles": { + "authored-output-contract-with-gallery-caption": "Original six-language output instructions around source gallery captions; captions are not claimed to be translated or natural requests", + "academic-text-answer": "Academic question requests a text answer, not image generation", + "aya-original-human": "Human-authored source requests, independently assistant-reviewed task labels", + }, + "image_task_policy": "AR includes text, code, source SVG/Mermaid, describing existing images; DIFFUSION requests a new image; BOTH explicitly requests a new image and a separate textual explanation", + "caption_counts": dict(counts), + } + print( + json.dumps(write_corpus(args.output, splits, contract, provenance)), flush=True + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/README.md b/reproducibility/src/training/model_classifier/sequence_repair/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0f8b911595a3219745081cb7ef932c0b821349ba --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/README.md @@ -0,0 +1,226 @@ +# Reproducible sequence-classifier repair + +This module initializes, continues, evaluates, and freezes standard Hugging Face +`ModernBertForSequenceClassification` checkpoints. Domain, FactCheck, and Modality +keep their existing label IDs. Safety, Feedback, and PromptGuard can reuse the +single-label runner; multi-label Hazard uses its own masked-loss training loop. +The shared loader and exporter preserve the contract's `problem_type`. + +The Vela family name uses the shared encoder scale, **307M**. Export receipts also +record the actual unique parameter count including each task head. A configured +32,768-position capacity is a limit, not evidence of long-document accuracy. + +## Data and provenance + +The [source manifest](data/source-files.json) pins public dataset revisions, +individual files, and SHA-256 values. It downloads metadata and text only, not +DiffusionDB images. Prepare one source directory: + +```bash +python -m src.training.model_classifier.sequence_repair.fetch_sources \ + --output artifacts/vela/sources +``` + +The reviewed-label sidecars contain source identifiers and task judgments rather +than copies of source requests. Hydration verifies the complete source file and +each request's text hash; Aya also verifies its contributor group. For example: + +```bash +python -m src.training.model_classifier.sequence_repair.hydrate_annotations \ + --sidecar src/training/model_classifier/sequence_repair/data/aya-reviewed-labels-v1.json \ + --source artifacts/vela/sources/CohereLabs--aya_dataset/data/train-00000-of-00001.parquet \ + --output artifacts/vela/annotations/aya-reviewed-labels-v1.jsonl +``` + +The same command reconstructs `aya-fact-reviewed-v2.json` and +`dolly-fact-reviewed-v2.json` using their specified source file. New task labels +were individually reviewed by an assistant before model predictions. Source +requests were human-authored; the new labels must not be described as expert +human annotations. Ambiguous requests and identified duplicate source/template +families retain their exclusion reasons. See [data provenance](data/README.md). + +Each task recipe writes `train.jsonl`, `dev.jsonl`, `test.jsonl`, `contract.json`, +and `manifest.json`. Rows contain `id`, `text`, `label`, and `group_id`; reported +metadata includes `source`, `language`, `length_bucket`, and `position`. IDs are +unique. Normalized duplicate text or source groups cannot cross partitions. +Missing/conflicting labels cause an error. Similarity screening and source-family +policy belong to the task recipe, not an inferred guarantee of this generic +validator. Old-weight and base-pretraining exposure may remain unknown. + +## Initialize and train + +The isolated reference environment used Torch 2.10 with ROCm 7, +Transformers 4.57.6, PEFT 0.18.1, and Accelerate 1.10.1. Data preparation additionally +uses PyArrow and scikit-learn. Pin the complete environment for a reproduced run; +do not install into a running inference service environment. + +For a fresh task head and LoRA adapter: + +```bash +python -m src.training.model_classifier.sequence_repair.initialize \ + --base artifacts/vela/base \ + --base-id llm-semantic-router/mmbert-32k-yarn \ + --base-revision 72a23a6640489471eb4ff7ad3ec5bc80af8a27de \ + --contract artifacts/vela/factcheck/contract.json \ + --output artifacts/vela/factcheck-initial +``` + +Initialization makes the standard head and classifier trainable in addition to +rank-32 LoRA on attention and MLP projections. For continuation, supply the selected +adapter explicitly instead. The published base configuration is retained; this +workflow does not infer or inject a missing RoPE-scaling setting from a model name. +An optional `classifier_pooling` contract field selects the standard HF `mean` +or `cls` mode. Omit it to retain the base's setting. Treat a pooling change as a +new trained candidate and compare short and long development quality; changing +the configuration alone does not make a mean-trained head a valid CLS model. + +```bash +python -m src.training.model_classifier.sequence_repair.train \ + --base artifacts/vela/base \ + --base-revision 72a23a6640489471eb4ff7ad3ec5bc80af8a27de \ + --adapter artifacts/vela/factcheck-initial \ + --contract artifacts/vela/factcheck/contract.json \ + --train artifacts/vela/factcheck/train.jsonl \ + --dev artifacts/vela/factcheck/dev.jsonl \ + --output artifacts/vela/runs/factcheck-short \ + --steps 600 --batch-size 8 --accumulate 2 --max-length 2048 \ + --learning-rate 0.00002 --eval-every 100 \ + --source-balanced-sampling --balanced-sampling --selection source-macro-f1 +``` + +The manual loop keeps parameters and cross-entropy accumulation in FP32, uses +BF16 autocast and SDPA, and enables non-reentrant gradient checkpointing. It +checks real losses and gradients for finiteness. Each equally sized microbatch +contributes mean CE divided by the number of accumulation steps. The optional +Torch test compares this gradient with a full batch; this avoids relying on +Trainer loss-kwargs behavior for a model that does not consume its normalization +argument. + +Selection evaluates the initial checkpoint as a quality floor, then uses only +development data. It writes `run.json`, per-step development metrics, +`selection.json`, `best-adapter`, and `last-adapter`. Existing runs are not +overwritten. The default score is macro F1 across the complete label ontology. +`source-macro-f1` averages that score equally across sources; `length-macro-f1` +averages it across length buckets. A source lacking a class still has that class +in the macro denominator, so inspect per-source accuracy and support as well. +The explicit `source-present-macro-f1` option instead averages only labels with +gold support within each source, then gives sources equal weight. It is useful +when an expanded taxonomy has sources that observe different label subsets. +Errors on observed labels still count, and all complete-ontology metrics remain +in the report. Existing selection modes and sampling are unchanged. +Source balancing samples a source first, optional length balancing then samples a +length, and optional label balancing then samples a label. These flags do not +copy rows or move groups between splits. An explicit `--source-weights` +JSON object instead controls source probabilities; it must name every source +remaining after context-budget filtering with a finite positive weight. It +implies source sampling and still permits length/label balance within each +source. For example, `--source-weights '{"natural": 9, "authored": 1}'` draws +approximately 90% from `natural`. Without this flag, original sampling and RNG +behavior are preserved. `sampling-step-*.json` and `sampling-final.json` record +per-source draws, unique-row coverage, and label draws. These receipts expose +repeated sampling of small sources; they are not quality metrics. + +### Optional token-budget microbatches + +`--microbatch-token-budget 32768` first samples the same +`batch-size × accumulate` examples using the existing RNG, stably orders those +examples by their measured length, and groups them so each padded microbatch +fits the token budget. Every example contributes CE sum divided by the original +optimizer-step example count. A 32K example therefore runs alone; shorter +examples can share a forward pass. Optimizer steps, learning-rate schedule, +sampling probabilities, and example coverage remain unchanged. The budget must +fit the longest eligible training example; no training row is silently dropped. + +This explicit option changes batching and the dropout random-number trajectory; +it is a new reproducible run, not byte-identical continuation of a prior run. +Dropout-disabled FP32 ModernBERT tests check gradient equivalence against one +complete batch. Steps record actual microbatch counts and maximum padded tokens; +measure speed on the intended workload. Without the option, the original +microbatch ordering and accumulation path remain unchanged. + +## Measure long context + +`prepare_context` creates exact attended-token budgets around frozen task +payloads. Backgrounds differ across train/dev/test and payload variants retain +their original groups. The supplied backgrounds are repeated, authored prose; +results are **context stress tests**, not estimates on natural long documents. +The label refers to the explicitly marked requested task. Report paired +length/position variants together rather than counting them as independent +natural requests. + +```bash +python -m src.training.model_classifier.sequence_repair.prepare_context \ + --corpus artifacts/vela/factcheck \ + --tokenizer artifacts/vela/base \ + --backgrounds src/training/model_classifier/sequence_repair/data/task-context-backgrounds.json \ + --output artifacts/vela/factcheck-context \ + --budgets 4096 8192 16384 32768 +``` + +Use a one-step probe on one actual long input before raising the training budget. +A long curriculum can pass both short and context files to `--train` and `--dev`, +with batch size 1, accumulation, and `--length-balanced-sampling`. Development +must expose short-quality regression as well as long-position failures. Inputs +over the explicit budget are never silently truncated; training rejections are +recorded and evaluation rejects an oversized input. + +## Freeze, then evaluate the independent test + +Some historical adapters save only `classifier`, leaving the sequence `head` +inherited from their original base. Before comparing that adapter on another +encoder base, preserve its original task head in a separate adapter artifact: + +```bash +python -m src.training.model_classifier.sequence_repair.complete_adapter \ + --base artifacts/vela/original-base \ + --adapter artifacts/vela/runs/domain-context/best-adapter \ + --contract artifacts/vela/domain/contract.json \ + --output artifacts/vela/domain-complete-head +``` + +This operation copies the existing head without training and checks CPU FP32 +logits before and after completion. It refuses adapters that already own `head`. +Use the same completed adapter for both sides of a base-migration comparison; +otherwise changing the base may also replace the task head. Select a base using +short and long development retention, and record the migration separately from +the original training receipt. A family name alone does not establish lineage. + +Export merges the selected adapter in FP32, checks short-input numerical +agreement before/after merging, preserves the standard architecture and label +mapping, and writes a `candidate-lock.json` with hashes. Supply the selected run's +receipt. The base revision must match that receipt and test selection must be +explicitly false. + +```bash +python -m src.training.model_classifier.sequence_repair.export \ + --runtime-task fact-check \ + --base artifacts/vela/base \ + --base-revision 72a23a6640489471eb4ff7ad3ec5bc80af8a27de \ + --adapter artifacts/vela/runs/factcheck-short/best-adapter \ + --run-manifest artifacts/vela/runs/factcheck-short/run.json \ + --contract artifacts/vela/factcheck/contract.json \ + --output artifacts/vela/frozen/factcheck + +python -m src.training.model_classifier.sequence_repair.evaluate \ + --base artifacts/vela/frozen/factcheck \ + --contract artifacts/vela/factcheck/contract.json \ + --data artifacts/vela/factcheck/test.jsonl \ + --output artifacts/vela/evidence/factcheck-final.json \ + --max-length 32768 --dtype bfloat16 +``` + +The required `--runtime-task` writes the router's actual sidecar schema alongside +the HF label mapping, before hashing the frozen candidate. Domain uses +`category_mapping.json`; FactCheck uses `fact_check_mapping.json`; PromptGuard +uses `jailbreak_type_mapping.json`; Modality also receives `modality_mapping.json` +and must retain the native `AR=0`, `DIFFUSION=1`, `BOTH=2` indices. Feedback, +Safety, and Hazard retain the complete named label set in `config.json` and +`label_mapping.json`. Every sidecar derives from the same classifier label IDs. + +Test is an explicit separate invocation, never run by training or export. Compare +baseline and candidate on the same rows, precision, and decoding policy; report +source/language/length breakdowns, uncertainty, and regressions. `--dtype +float32` disables GPU autocast for a precision-controlled comparison. Forward +batch timings exclude tokenization and HTTP serving overhead and are not request +latency. Native and ONNX execution need their own parity and serving validation. +A short merge-equivalence check alone is not a quality or 32K-runtime test. diff --git a/reproducibility/src/training/model_classifier/sequence_repair/__init__.py b/reproducibility/src/training/model_classifier/sequence_repair/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..f96c02f070be5df256e3d31f760f338b43b28adc --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/__init__.py @@ -0,0 +1 @@ +"""Reproducible continuation of fixed-label sequence classifiers.""" diff --git a/reproducibility/src/training/model_classifier/sequence_repair/complete_adapter.py b/reproducibility/src/training/model_classifier/sequence_repair/complete_adapter.py new file mode 100644 index 0000000000000000000000000000000000000000..2bd12e5c2527e9e71ddd5864c5def76bd6c9b9b6 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/complete_adapter.py @@ -0,0 +1,89 @@ +"""Make a legacy sequence adapter self-contained with its original task head.""" + +# ruff: noqa: PLC0415 +import argparse +import hashlib +import json +import shutil +from pathlib import Path + +from .model import load_model + + +def main(): + import torch + from safetensors.torch import load_file, save_file + + parser = argparse.ArgumentParser() + parser.add_argument("--base", required=True) + parser.add_argument("--adapter", type=Path, required=True) + parser.add_argument("--contract", required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + if args.output.exists(): + raise ValueError("Refusing to overwrite a completed adapter") + config = json.loads((args.adapter / "adapter_config.json").read_text()) + if "head" in (config.get("modules_to_save") or []): + raise ValueError("Adapter already owns its task head") + torch.set_num_threads(8) + model, tokenizer, _labels, _ids = load_model(args.base, args.contract, args.adapter) + model.eval() + tokens = tokenizer( + ["A task head parity probe.", "核对分类头参数是否保持不变。"], + padding=True, + return_tensors="pt", + ) + with torch.inference_mode(): + before = model(**tokens).logits + head = model.get_base_model().head.state_dict() + weights = load_file(args.adapter / "adapter_model.safetensors", device="cpu") + for name, tensor in head.items(): + key = "base_model.model.head." + name + if key in weights: + raise ValueError("Unexpected existing head tensor in legacy adapter") + weights[key] = tensor.detach().cpu().contiguous() + args.output.mkdir(parents=True) + for path in args.adapter.iterdir(): + if path.is_file() and path.name not in { + "adapter_config.json", + "adapter_model.safetensors", + }: + shutil.copy2(path, args.output / path.name) + config["modules_to_save"] = [*(config.get("modules_to_save") or []), "head"] + (args.output / "adapter_config.json").write_text( + json.dumps(config, indent=2) + "\n" + ) + save_file( + weights, args.output / "adapter_model.safetensors", metadata={"format": "pt"} + ) + restored, _tokenizer, _labels, _ids = load_model( + args.base, args.contract, args.output + ) + restored.eval() + with torch.inference_mode(): + after = restored(**tokens).logits + if not bool(torch.isfinite(before).all() and torch.isfinite(after).all()): + raise ValueError("Task-head completion produced non-finite logits") + torch.testing.assert_close(before, after, rtol=1e-6, atol=1e-6) + receipt = { + "operation": "Preserve the existing sequence task head in a legacy adapter; no training or encoder change", + "source_adapter_sha256": hashlib.sha256( + (args.adapter / "adapter_model.safetensors").read_bytes() + ).hexdigest(), + "completed_adapter_sha256": hashlib.sha256( + (args.output / "adapter_model.safetensors").read_bytes() + ).hexdigest(), + "base_model": config["base_model_name_or_path"], + "base_revision": config.get("revision"), + "head_tensors": list(head), + "maximum_absolute_logit_difference": float((before - after).abs().max()), + "quality_evaluation": False, + } + (args.output / "task-head-completion.json").write_text( + json.dumps(receipt, indent=2) + "\n" + ) + print(json.dumps(receipt)) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/context.py b/reproducibility/src/training/model_classifier/sequence_repair/context.py new file mode 100644 index 0000000000000000000000000000000000000000..e4b1ae27b178e19cb770cf1e5d98557e0fa27895 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/context.py @@ -0,0 +1,85 @@ +"""Construct measured context probes without truncating the task-bearing payload.""" + + +def insert_payload( + payload, + background, + position, + budget, + count_tokens, + *, + fill_to_budget=True, + neutral_suffix=" note" +): + if position not in {"head", "middle", "tail"}: + raise ValueError("Unknown payload position") + if not payload or not background.strip() or budget <= 0: + raise ValueError("Payload, background and token budget must be nonempty") + if count_tokens(payload) > budget: + raise ValueError("Payload itself exceeds the token budget") + if count_tokens(background) <= count_tokens(""): + raise ValueError("Background must contribute actual tokens") + + def build(repetitions): + left = {"head": 0, "middle": repetitions // 2, "tail": repetitions}[position] + prefix = "\n".join([background] * left) + suffix = "\n".join([background] * (repetitions - left)) + text = ( + (prefix + "\n" if prefix else "") + + payload + + ("\n" + suffix if suffix else "") + ) + start = len(prefix) + 1 if prefix else 0 + return text, start + + measured = {} + + def measure(repetitions): + if repetitions not in measured: + measured[repetitions] = count_tokens(build(repetitions)[0]) + return measured[repetitions] + + # Repeated complete paragraphs usually have a constant token increment. + # Use that only to find a close starting point; exact measurements still + # bracket and determine the result, including boundary-token differences. + increment = max(1, measure(2) - measure(1)) + estimate = min(budget, max(0, 1 + (budget - measure(1)) // increment)) + if measure(estimate) <= budget: + low, high, distance = estimate, estimate, 1 + while high < budget: + high = min(budget, estimate + distance) + if measure(high) > budget: + break + low, distance = high, distance * 2 + else: + low, high, distance = estimate, estimate, 1 + while low > 0: + low = max(0, estimate - distance) + if measure(low) <= budget: + break + high, distance = low, distance * 2 + while low < high: + middle = (low + high + 1) // 2 + if measure(middle) <= budget: + low = middle + else: + high = middle - 1 + text, start = build(low) + actual = measure(low) + if fill_to_budget: + text += neutral_suffix * (budget - actual) + actual = count_tokens(text) + if actual != budget: + raise ValueError( + "Neutral suffix does not fill this tokenizer's exact budget" + ) + if actual > budget or text[start : start + len(payload)] != payload: + raise ValueError( + "Context construction corrupted its payload or exceeded budget" + ) + return { + "text": text, + "payload_start": start, + "payload_end": start + len(payload), + "actual_tokens": actual, + } diff --git a/reproducibility/src/training/model_classifier/sequence_repair/corpus.py b/reproducibility/src/training/model_classifier/sequence_repair/corpus.py new file mode 100644 index 0000000000000000000000000000000000000000..1497948584e6613d312c94bb0d259c14c22c6b3b --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/corpus.py @@ -0,0 +1,55 @@ +"""Frozen task corpus receipts and source-group partitioning.""" + +import hashlib +import json +from pathlib import Path + +from .data import assert_disjoint, normalized_text + + +def group_split(group, seed=20260913): + bucket = int(hashlib.sha256(f"{seed}:{group}".encode()).hexdigest()[:8], 16) % 100 + return "train" if bucket < 70 else "dev" if bucket < 85 else "test" # noqa: PLR2004 + + +def write_corpus(output, splits, contract, provenance): + output = Path(output) + if output.exists() and any(output.iterdir()): + raise ValueError("Refusing to overwrite a frozen corpus") + unique = {} + for split, rows in splits.items(): + seen, retained, retained_keys = {}, [], set() + for row in rows: + key = normalized_text(row["text"]) + previous = seen.setdefault(key, row["label"]) + if previous != row["label"]: + raise ValueError("Contradictory labels for identical normalized text") + if key in unique and unique[key] != split: + raise ValueError("Duplicate text crosses corpus splits") + unique[key] = split + if key not in retained_keys: + retained.append(row) + retained_keys.add(key) + splits[split] = retained + for first, second in [("train", "dev"), ("train", "test"), ("dev", "test")]: + assert_disjoint(splits[first], splits[second]) + output.mkdir(parents=True, exist_ok=True) + files = {} + for split, rows in splits.items(): + content = "".join( + json.dumps(row, ensure_ascii=False, sort_keys=True) + "\n" for row in rows + ) + (output / f"{split}.jsonl").write_text(content) + files[split] = { + "rows": len(rows), + "sha256": hashlib.sha256(content.encode()).hexdigest(), + } + (output / "contract.json").write_text(json.dumps(contract, indent=2) + "\n") + manifest = { + "files": files, + "provenance": provenance, + "test_used_for_selection": False, + "split_policy": "Group-disjoint before sampling; exact NFKC/case/whitespace overlap rejected", + } + (output / "manifest.json").write_text(json.dumps(manifest, indent=2) + "\n") + return manifest diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data.py b/reproducibility/src/training/model_classifier/sequence_repair/data.py new file mode 100644 index 0000000000000000000000000000000000000000..eace1fb7191782f579f221baa782c71e82e886fc --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data.py @@ -0,0 +1,107 @@ +"""Dependency-light sequence data contracts and classification metrics.""" + +import hashlib +import json +import unicodedata +from collections import Counter +from pathlib import Path + + +def normalized_text(text): + return " ".join(unicodedata.normalize("NFKC", text).casefold().split()) + + +def load_contract(path): + config = json.loads(Path(path).read_text()) + id_to_label = {int(index): label for index, label in config["id2label"].items()} + if set(id_to_label) != set(range(len(id_to_label))): + raise ValueError("Label IDs must be contiguous from zero") + if config["label2id"] != {label: index for index, label in id_to_label.items()}: + raise ValueError("id2label and label2id must be a bijection") + return config["label2id"], id_to_label + + +def read_records(paths, label_to_id): + records, ids, texts = [], set(), {} + for path in paths: + for line in Path(path).read_text().split("\n"): + if not line.strip(): + continue + row = json.loads(line) + if not isinstance(row.get("text"), str) or not row["text"].strip(): + raise ValueError("A classification row needs nonempty text") + if not row.get("id") or row["id"] in ids: + raise ValueError("Missing or duplicate row ID") + if not row.get("group_id"): + raise ValueError("Each row needs a source/translation/template group") + if row.get("label") not in label_to_id: + raise ValueError("Unknown classification label") + normalized = normalized_text(row["text"]) + previous = texts.setdefault(normalized, row["label"]) + if previous != row["label"]: + raise ValueError("Identical text has contradictory labels") + ids.add(row["id"]) + records.append(row) + if not records: + raise ValueError( + "Empty datasets cannot produce training or evaluation evidence" + ) + return records + + +def assert_disjoint(train, dev): + for key, transform in [("group_id", str), ("text", normalized_text), ("id", str)]: + overlap = {transform(row[key]) for row in train} & { + transform(row[key]) for row in dev + } + if overlap: + raise ValueError(f"Training/development overlap in {key}: {len(overlap)}") + + +def file_receipts(paths): + return [ + { + "file": Path(path).name, + "sha256": hashlib.sha256(Path(path).read_bytes()).hexdigest(), + } + for path in paths + ] + + +def classification_metrics(gold, predicted, id_to_label): + if not gold or len(gold) != len(predicted): + raise ValueError("Empty or mismatched prediction evidence") + count = len(id_to_label) + confusion = [[0] * count for _ in range(count)] + for index, expected in enumerate(gold): + actual = predicted[index] + if expected not in id_to_label or actual not in id_to_label: + raise ValueError("Prediction outside the frozen label contract") + confusion[expected][actual] += 1 + per_label = {} + for index, label in id_to_label.items(): + tp = confusion[index][index] + support = sum(confusion[index]) + selected = sum(row[index] for row in confusion) + per_label[label] = { + "support": support, + "tp": tp, + "fp": selected - tp, + "fn": support - tp, + "precision": tp / selected if selected else 0.0, + "recall": tp / support if support else 0.0, + "f1": 2 * tp / (selected + support) if selected + support else 0.0, + } + present = [row for row in per_label.values() if row["support"]] + return { + "rows": len(gold), + "accuracy": sum(confusion[index][index] for index in range(count)) / len(gold), + "macro_f1": sum(row["f1"] for row in per_label.values()) / count, + "macro_f1_present_labels": sum(row["f1"] for row in present) / len(present), + "per_label": per_label, + "confusion": confusion, + } + + +def label_counts(records): + return dict(Counter(row["label"] for row in records)) diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/README.md b/reproducibility/src/training/model_classifier/sequence_repair/data/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0a99f0e62a59f5ef8a6e6a2ee08405b801cb0cbb --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/README.md @@ -0,0 +1,49 @@ +# Reviewed task-data provenance + +`source-files.json` pins the exact public text/metadata files used by the recipes. +Paths use repository IDs with `/` replaced by `--`. The MMLU-Pro and category +supplement files are historical-exposure screening inputs for Domain; they are +not claimed to be a new independent test set. + +- Aya: `CohereLabs/aya_dataset`, revision + `f9ea04583f02a8f86404ff6c58bf75fe637df8a2`, Apache-2.0. +- Global-MMLU: `CohereLabs/Global-MMLU`, revision + `0e619dbeb34206cd48705a1a0ea7fb21cae09993`, Apache-2.0. +- DiffusionDB captions: `poloclub/diffusiondb`, revision + `fb620fbe49fa4420e0734bd9c0df11f51176b61f`, CC0-1.0. No images are used. +- Dolly: `databricks/databricks-dolly-15k`, revision + `bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a`, CC-BY-SA-3.0. + Copyright 2023 Databricks, Inc.; Wikipedia editors and contributors for source + passages. Retain source attribution and applicable license terms when + distributing reconstructed data or derived annotations. The + [official card](https://huggingface.co/datasets/databricks/databricks-dolly-15k/blob/bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a/README.md) + describes the employee-written instructions and Wikipedia context fields. + +The three reviewed sidecars contain 216 original Aya requests with per-task +labels, 59 additional Aya FactCheck requests (57 retained), and 192 Dolly +FactCheck requests (177 retained). Every task judgment was made by an assistant +reading the request before model predictions; these labels are not human-expert +annotations. Excluded records preserve the review reason. Source category does +not assign the FactCheck label: for example, a Dolly brainstorming request can +still ask for factual country names, while an extraction request can be answered +entirely from its supplied context. + +Aya contributors form split groups. Dolly uses the source passage, or instruction +when no passage exists, as its grouping key. Identified duplicate passage topics +and instruction-template families across partitions are excluded in the reviewed +set. These checks cannot establish that old published weights or the base +encoder never saw the source data. Earlier FactCheck recipes used some Dolly +instructions and did not preserve a complete row-level training manifest. + +`hydrate_annotations` verifies the source-file hash and reconstructs the exact +reviewed text. Reconstructed v2 JSONL checksums are +`140e6f04b9ec52dd9dfaac2322f6180ffc0c292859a43d0423557940b48c7df5` +for Aya and +`b8c1a2e3badef640f08ee1a68240fed7647d0902bd4fabf833a99008c2d17cd0` +for Dolly. Both were reproduced against the pinned source files. + +`task-context-backgrounds.json` contains original prose for controlled context +stress. Train, development, and test use different background narratives. +Paragraphs repeat within an input to reach measured token budgets. A source +request reused at several lengths or positions is a paired variant, not a new +independent natural example. diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/aya-fact-reviewed-v2.json b/reproducibility/src/training/model_classifier/sequence_repair/data/aya-fact-reviewed-v2.json new file mode 100644 index 0000000000000000000000000000000000000000..58a89a62804120ca3c62f4ec6fc43996153df2b8 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/aya-fact-reviewed-v2.json @@ -0,0 +1,840 @@ +{ + "source": { + "repository": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "filename": "data/train-00000-of-00001.parquet", + "file_sha256": "51baa85043b569ba117f41516b32d1b0e4e2647fb203a08933f98a556fb1fb06", + "license": "Apache-2.0" + }, + "annotation_scope": "Individually assistant-reviewed task labels on human-authored source requests, assigned before model predictions. These are not human-expert task annotations. Excluded ambiguous requests are retained with reasons.", + "records": [ + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-0", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "342524f7e8a304b3e054f4cab18abd77a5ef1a69f1676b9e6dc43e2d6b65b88c", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-79ce9989c35133a477f1963517517015315ffdfa8955012a1d0809882c919403", + "id": "fact-natural-v2-eng-dev-1", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8f0b3389f8ff0f27db0e6642cbb83960515d05f4f3efdea250249094833ceffd", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-2", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "a7cd916823157509713a323136030257d7f2d9ed56f6fcf8b9ef291d2539b6b5", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-3", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "a7711471ca026631a3d9e8a576b3d14a8f0d28679b86866ac1bba61bb51fdb1e", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-ffded8291f05fcba12aa6cdc6fbe8edb3b17a2176879746f2acc29e6c4b395bd", + "id": "fact-natural-v2-eng-dev-4", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "332fcc50f6565582768e5a8ce48eb4890332fce397dc80523a171145ed4940a4", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-5", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "031ee4ff7d629172f6d1fd14b6f8651b365058ba7b62579b6175c14515703d18", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-6", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "5afa00014668fd3a6c8765377d9c00553bbf4fc2599b37a19866098b3a97a323", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "id": "fact-natural-v2-eng-dev-7", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "5f8f9cc2d1f8a3ae64b920491e16e5730ff146c6e6d70a5d09f6b0a8b373864b", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-637eec2fa2066e814aefcd7e0ef0cd4d5e61168e06a75a4e0360bab5868b7ed8", + "id": "fact-natural-v2-eng-test-0", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "f4fba18f4454b334ed2c2b179a642d71ffe2093df86a64d385ca5fc975d4275c", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", + "id": "fact-natural-v2-eng-test-1", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8332918127dd8f661c913f3a1a21c914f18ea27203e020e45bac1994c12df4ad", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-37202666dac92f07d1a908cf26159245f069fcc2b14346812d45fa754a28a7a7", + "id": "fact-natural-v2-eng-test-2", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "1f081fc1a6874e457528f5334d7ca5639d7ed96b626332217d98f1a9bcc4b011", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-37202666dac92f07d1a908cf26159245f069fcc2b14346812d45fa754a28a7a7", + "id": "fact-natural-v2-eng-test-3", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "b969d7d8ffcbc90922fb057b437315853c0463e9cc0f25c8e353a0c50bb8b88a", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", + "id": "fact-natural-v2-eng-test-4", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "c8438f3098c2ff6f26a09edcef9e217dd4c13bfaf1d8cf6330db2b91dd8e23e1", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", + "id": "fact-natural-v2-eng-test-5", + "label": "NO_FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "d38321b60249f4b1191ea4295a920802bd2b417791e21d226cf458ddab567ba4", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-3ea821a41f2e7d0bd1af5f2f52f0548bcdb19b08c65001bb450319472007a8c2", + "id": "fact-natural-v2-eng-test-6", + "label": "FACT_CHECK_NEEDED", + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8a172dffba9531ac2deaf25e68fa396950922458dbe5b4d46dbe43409b5fc8f3", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "exclusion_reason": "Incomplete request: only a word without requested operation", + "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", + "id": "fact-natural-v2-eng-test-7", + "label": null, + "language": "eng", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "116c177309716229b58ac2cec516d028b615ba8acecb66da533d5fb1f183b6b9", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-0", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "bebdff2ba53e17109b203134cfd4df53260485e09667844dd8ebb6a8adc42acf", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-1", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "d4f1a34a2f173fc82e1c64142b4c38812253d0eb7059e3d6834e67bcaa5090c9", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-2", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "377511335934ee85fd145d638120116686c16a414a0c0580ef81a04cc9a45a77", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-3", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "396a485e330e0439c3b92a373e7b3d81618096e68f6d8dd4d33eb8a24e0fa2ca", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-4", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "d114806e609149fa5583c2b727d69384dc96ea1ea0f790a6c5a9ac1d85763274", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-5", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "9a1bce44fd74cb42ea60e2ea88db140fc822a93b78e2b3f1c896adc25f5a29eb", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-6", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "c9e7d288c0c42f54d6bf51d52ce090adc5815b5d4522c08aed35153522cf733f", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "id": "fact-natural-v2-zho-dev-7", + "label": "FACT_CHECK_NEEDED", + "language": "zho", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "b6fb1fb1df5145031645cb0abfd47696955da4ec44ff75e3dae5915154b46f00", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "id": "fact-natural-v2-spa-dev-0", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "c7a51d11d6700354f4668e56392acc000cf38fb6cd776f1bc26b1428903144b4", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "id": "fact-natural-v2-spa-dev-1", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "bca45f2318687f31927c7c4184afda91ab9582f88e8a425b9a788e7b70b413f7", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "id": "fact-natural-v2-spa-dev-2", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "90c972fabe42f248af8574eeae446af0f241ffd8069cbffe8f487ba241388bcd", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-8b952b20ac15bab5342d070620109cbbcd07edb78f7c7ae9b655edfc8f3c1f53", + "id": "fact-natural-v2-spa-dev-3", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8da1d36edeb5d015eaea7ab947ef49be3d8949fa2727534f716bb046871495f1", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "id": "fact-natural-v2-spa-dev-4", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "511d865af1b30a6f6ceaa45000e44195676f0002f92366c620e58516ebf00b1b", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "id": "fact-natural-v2-spa-dev-5", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "eb709d1f9fdded428a4117bcf1f5588de0de3fca72cef691673564c174bacea6", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "id": "fact-natural-v2-spa-dev-6", + "label": "NO_FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "93fcb0d58ef78fbf215ea7d362275e394d6dd38d53763b0f4ea004f22ae97552", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-8b952b20ac15bab5342d070620109cbbcd07edb78f7c7ae9b655edfc8f3c1f53", + "id": "fact-natural-v2-spa-dev-7", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "50104a465dc4ea6f9abd7236adc1a8073077dfabfe43ced678b73e26540cd407", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-0", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "7f919debf88633f014530977dce4f525e89606fd7a780b59e57338a445e5dd24", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-1", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8b2e53da6a58560f8b088522c0d2eb9181671e8658d3d1973943dfe5b07e3b4c", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-2", + "label": "NO_FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "4a3cf9040d9108b5b444d248586aaee3d7ce455cba16072f5ecbd55af989f9bd", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-3", + "label": "NO_FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "17fc5c8d9d5931cdb6f0e6dabf2126f7403e7ee46c96a721bc5d320ffb01fe10", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-4", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "962c3731a64f0e9f8059854eb975fff03916c633c49426b9dfe13c8b82715532", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-5", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "e247d518c4f2e4f93232f343f2e2d96d9dc8515b33d5146ef4131a128ba6f091", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "id": "fact-natural-v2-spa-test-6", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "abbb49198bcb4abb83f562921063694bf5c28b265fb197361598da15d41763cb", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-2986e8162ec136cdcfdd5402e3765f8fa028d4f1726dfec5e501a6c650f5adbf", + "id": "fact-natural-v2-spa-test-7", + "label": "FACT_CHECK_NEEDED", + "language": "spa", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "724bca7cbc0ce46b5e16e71438d8460392fe74bd0ccffb3cd711d8f1e2e25a61", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-0", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "task-transform-candidate", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "b808b9fba09d787837e369d03e0a515a2ad83ac24bf42b7085fd43c4f20b0281", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-1", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "task-transform-candidate", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "a9ddef6da6b470409773c2830a2096d7ae327b0738f6c9084071807e8747091b", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-2", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "20a6719209b9deef6a16a10074bbd62e33722ef099f8e4a2298e331f9af2d206", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-3", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "fbe302a6ad95a8ba7c9f28d050b43ef33b716bf13dbc5eb99a536e9e85d8219e", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "exclusion_reason": "Unspecified competitions, subject cannot be resolved", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-4", + "label": null, + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "c47957489831d3c97d4ab9c669d5b6eb219aad3ff7c8bdcd194dbe4aec8e53d9", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-5", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "137336e74c681cbd5077b48232438a4370f7f495ba931ebf684a91134f0ef3e5", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-6", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "142122c3434772068db998af78602ef7aa4ba224f294d4aa06a51fd8c7e972d8", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-7", + "label": "NO_FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "9a598fbbbd4fd4d87ab538aacd7313d025ccab51b49177dedf3db47be5f432fe", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-8", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "e0b962ddabd607b35f267b18305aa11993f506a6396c24bde375e672b758acb1", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "id": "fact-natural-v2-fra-test-9", + "label": "FACT_CHECK_NEEDED", + "language": "fra", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "9199cb4b442bc08cfe4749a743834b5853d0c25d35572ba0374acb10abb49218", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-08cf632b35df86fe1b539fa75107c5c6c9a73f4a7a912cf5c97bc0a75838e211", + "id": "fact-natural-v2-deu-dev-0", + "label": "FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "6de763142163a1e3bdbc58af771f5cca4f5d0ed109fcd4f09de5d4d705a38d70", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-08cf632b35df86fe1b539fa75107c5c6c9a73f4a7a912cf5c97bc0a75838e211", + "id": "fact-natural-v2-deu-dev-1", + "label": "FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "f258206774b964bd378f074c1eba2dc2d6024ac74e235e5af51e6e3564b8b852", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-89da884608fffca31bdfae1e82eef0f2159a3cf0c47f6d6ffe01f25c6aa6bb42", + "id": "fact-natural-v2-deu-dev-2", + "label": "FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "aa56c0379a194c946c139c81a817e854d1c11f5cd74079e3f838e8dc25ebcfe9", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-08cf632b35df86fe1b539fa75107c5c6c9a73f4a7a912cf5c97bc0a75838e211", + "id": "fact-natural-v2-deu-dev-3", + "label": "NO_FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "fc034c0bf3e26d203542b024efaff4cc1b79c63e6afa922db5887c955239cc46", + "split": "dev" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "id": "fact-natural-v2-deu-test-0", + "label": "FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "8e6d1f2a08928811de61b2aa60ab8b0bd02c751370abde89830596d54b759d66", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "id": "fact-natural-v2-deu-test-1", + "label": "FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Answer requires externally grounded factual knowledge or verification", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "9afd9b3650aac9aa0b0e0f277f140ada3e57e5fd766625cf0386af1e1e38958b", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "id": "fact-natural-v2-deu-test-2", + "label": "NO_FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "a6fe736c225555776261b70609849e419e2e12992843b3a4e5734497381d167d", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "id": "fact-natural-v2-deu-test-3", + "label": "NO_FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "f0c2d35055a349fb996cbdf73eb6b24259c1110c7031e134d0f619f30fddec97", + "split": "test" + }, + { + "annotation_method": "Individually read original request before predictions; no topic/category label mapping", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "id": "fact-natural-v2-deu-test-4", + "label": "NO_FACT_CHECK_NEEDED", + "language": "deu", + "length_bucket": "short", + "review_rationale": "Self-contained extraction, linguistic transformation, subjective/creative or procedural task", + "reviewer": "assistant", + "sampling_stratum": "random-natural", + "source": "aya-original-human-reviewed-v2", + "source_text_sha256": "7450ab9f59dd4005c63eff9ba7691bc40f80ddff68eb473f167a12ff85afeb31", + "split": "test" + } + ] +} diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/aya-reviewed-labels-v1.json b/reproducibility/src/training/model_classifier/sequence_repair/data/aya-reviewed-labels-v1.json new file mode 100644 index 0000000000000000000000000000000000000000..bc19bbde40c8259c9af8c7720e6a911c6e883114 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/aya-reviewed-labels-v1.json @@ -0,0 +1,2839 @@ +{ + "source": { + "repository": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "filename": "data/train-00000-of-00001.parquet", + "file_sha256": "51baa85043b569ba117f41516b32d1b0e4e2647fb203a08933f98a556fb1fb06", + "license": "Apache-2.0" + }, + "annotation_scope": "Individually assistant-reviewed task labels on human-authored source requests, assigned before model predictions. These are not human-expert task annotations. Excluded ambiguous requests are retained with reasons.", + "records": [ + { + "id": "aya-review-eng-0", + "language": "eng", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "source": "aya-original-human", + "source_text_sha256": "00430cae5a93274e629db008f8ae1e8e78d164784098eed1314b4964939e8ba6", + "task_labels": { + "domain": "psychology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-1", + "language": "eng", + "group_id": "aya-author-aa8d2326c36fb9d28cb1d8aaab266ccb7ec8ffee879619612d22791b5fb4e61c", + "source": "aya-original-human", + "source_text_sha256": "0044b19fb316e9609f82522c4329615a1ef91f81c5fdf0c90e6af1d57fba8872", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-2", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "0058bcdb1ce6d8f4ee7a2adfa487ec35fc62b7d6f799fa4902416d19e24a8247", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-3", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "006310b9ce7f751e1544e9e26cea602e62ab6117a255b0f5cba4d19062902e5d", + "task_labels": { + "domain": "business", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-4", + "language": "eng", + "group_id": "aya-author-ca23f2d689e06c12d3ad3f5e33c7ca9c970b0fbb83b1cd49e2f80972794fc109", + "source": "aya-original-human", + "source_text_sha256": "00a8540a00bc4e10bd6fae0cfd0a2d6cdb6311675ea7225cad4525b552d995de", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-5", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "00e575522f85140b44ff15d08d294e759027e7fdd8d3fe90b20768b6f0a1e192", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-6", + "language": "eng", + "group_id": "aya-author-5c39b84813508cee899929f557fea6b9fda8282fd937c6620c4b70de8a7e93ab", + "source": "aya-original-human", + "source_text_sha256": "0109b434b0e153a55eb38bef330d5fa8a77aa2a06ad5284d16d795139c581098", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-7", + "language": "eng", + "group_id": "aya-author-5c39b84813508cee899929f557fea6b9fda8282fd937c6620c4b70de8a7e93ab", + "source": "aya-original-human", + "source_text_sha256": "013538bf460bb70e29f5bf19ed7645987943b5cdec3a2763f5d570342dfe1fb5", + "task_labels": { + "domain": "economics", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-8", + "language": "eng", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "source": "aya-original-human", + "source_text_sha256": "015a499f30a0ad978440041c657f3872513467b66aacc7427076370717e7849f", + "task_labels": { + "domain": null, + "factcheck": null, + "modality": null + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-9", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "017a8622da2c61de9ea9f67c0a6b221173c447e90b1fc4622b35961a239ef132", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-10", + "language": "eng", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "source": "aya-original-human", + "source_text_sha256": "01cbfa3ef5cc5c7b869894ea3fcf88a58e1d624950443971f47604447bcba052", + "task_labels": { + "domain": "psychology", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-11", + "language": "eng", + "group_id": "aya-author-59ad3805e07eacf4acbe528b017f4b114a2aad3449fc4ab27c167faf5ba16f18", + "source": "aya-original-human", + "source_text_sha256": "0236291d8536e4da67138c0291d4a2ee41f66b69d350ba2e190201cf6f2b613c", + "task_labels": { + "domain": "math", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-12", + "language": "eng", + "group_id": "aya-author-59ad3805e07eacf4acbe528b017f4b114a2aad3449fc4ab27c167faf5ba16f18", + "source": "aya-original-human", + "source_text_sha256": "0256add2358852ea8378cba81b94e21bbe53a626335a85e6f1ba5dd6b1555654", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-13", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "02929fda47428f2b4d9ec54187405fd6e03b0cb0eb99c392a6aa9d8d7ee4a811", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-14", + "language": "eng", + "group_id": "aya-author-8a7a3d8dc5a2c4d7fcccf525725a05889e9ebfa97816aa13721d9ec7c20bfdbe", + "source": "aya-original-human", + "source_text_sha256": "02999a8e70e1c28b80e2bc80e22db369b92ec99cb1daa1a5b2a08457f232d529", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-15", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "02e882b7a3ae29ddfb2bae72b75b402d24ae55d633f3184f0b376b72d8828290", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-16", + "language": "eng", + "group_id": "aya-author-5c39b84813508cee899929f557fea6b9fda8282fd937c6620c4b70de8a7e93ab", + "source": "aya-original-human", + "source_text_sha256": "02ee0f8bc2a34c6ce75a0410cf0c702dff881d3135b79174c0131349e4d72829", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-17", + "language": "eng", + "group_id": "aya-author-8c8ad332c18e58c4c2fdd2e1681ead0b0ee49d6fe446961226924a412856dac3", + "source": "aya-original-human", + "source_text_sha256": "0300747d2056b99b7d0d966c6d78ae8b0ecec93418c3fdd0bfc1b9a45acf2c9b", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-18", + "language": "eng", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "source": "aya-original-human", + "source_text_sha256": "031ee4ff7d629172f6d1fd14b6f8651b365058ba7b62579b6175c14515703d18", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-19", + "language": "eng", + "group_id": "aya-author-9e170d910e2194ebf3bb56437ea9d419733d407cecbfdddef1490d919ae7a15f", + "source": "aya-original-human", + "source_text_sha256": "034a3c0677e86cf11466746d0c05b5d44a8351d17347ae53efa86f7e35f776e1", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-20", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "037378d6ea1b234503b9926215acfd0349f7bd124fbc87da76b397cc6d309f9d", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-21", + "language": "eng", + "group_id": "aya-author-5c39b84813508cee899929f557fea6b9fda8282fd937c6620c4b70de8a7e93ab", + "source": "aya-original-human", + "source_text_sha256": "03771d4958e00b5c6c03f37926927d2363bc889f8f599ce6627d00755ff5f7a3", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-22", + "language": "eng", + "group_id": "aya-author-c9f8d0c12f63201fcb3216626ce4af84e873e2908bab98a1386f2c817de4ef15", + "source": "aya-original-human", + "source_text_sha256": "0381b69225f721023a920cde04af78120404456256b899546779d3bb8207a0a1", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-23", + "language": "eng", + "group_id": "aya-author-5bc7b2fd70f142b113cee411e3476425d7c987c0591da2752da99cc915cb14ae", + "source": "aya-original-human", + "source_text_sha256": "03838ea8898f133efd928a4ed679e5a0cb1fca9434809bfc7c548b0d94279ab0", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-24", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "038623de5d6a65a3a223a3c4fe68c523582edf66e85a0a8dee860ea6240b2067", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-25", + "language": "eng", + "group_id": "aya-author-924d74be8d40510a070fdf4ae61e1ff38a1719f39bba1ea928b41b264bf7c464", + "source": "aya-original-human", + "source_text_sha256": "03a429d177678beeffa7279add54a28cf673111b8e865981b46cca9a0c9d596b", + "task_labels": { + "domain": null, + "factcheck": null, + "modality": null + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-eng-26", + "language": "eng", + "group_id": "aya-author-6978e3ae49981e2ac48d57635b67aeea276c633bbe82d88f4411aa212e319372", + "source": "aya-original-human", + "source_text_sha256": "03da286646085682a776eec98bec1f6276c01eeb16b7e79b63dca8b34ec3a67d", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-27", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "044f790dc8608c2382c93962e2ad7f0d7860888c58680cda586cfc39dc538e32", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-28", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "049ca7bcee40a632cbeac67144f794780621349ea6076f9bc0e1556588f2bb22", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-29", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "04ae645fc4c88cae00598d7c2ffd390e8892ddcdd33f3654b4e650ee26baea07", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-30", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "0582af6d8dffc08fb47012a569075696250b5eecda2ef1951ae10859c2786aba", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-31", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "0586d6d6f291504cafb8fbd8b6e0f405603c2ef84c348885e2fe6fb8c47a635b", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-32", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "05b05aafb8d9fe843d57d13768a055ca8a0e7d7dab41c5180dc6640720a46539", + "task_labels": { + "domain": "math", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-33", + "language": "eng", + "group_id": "aya-author-2816aa9162735bfc3c72b16925f9527c78fedeab522f067276a78ccae0c1e2c1", + "source": "aya-original-human", + "source_text_sha256": "05be7a9c0ff2a991e345847288f2f5490ea435560e1c440d90703b620588549e", + "task_labels": { + "domain": "health", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-34", + "language": "eng", + "group_id": "aya-author-5bc7b2fd70f142b113cee411e3476425d7c987c0591da2752da99cc915cb14ae", + "source": "aya-original-human", + "source_text_sha256": "05f28ccad00db971734c8db12cb883dfaab064d892dce2b7a182587ef2a5ba4a", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-eng-35", + "language": "eng", + "group_id": "aya-author-e98d4e0d6460422079369572b448ae8cbad51db5c6f4e6451a5f065fb3deacea", + "source": "aya-original-human", + "source_text_sha256": "063530373bf9e0c03a8b104ed9338a3f8b36e19af4867623cf40a29ebc24a91a", + "task_labels": { + "domain": "math", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-0", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "001cce84daf06e862d2dbe7ddc4527bd17ea5ec68e80645ccd711f3d939cdbce", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-1", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "00388d0be5eafbd5ebd60ad2147b5f0cb31aeeab429f474a1c210c4b0accd80b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-2", + "language": "zho", + "group_id": "aya-author-68f2ab39801cc0fce2d39ceedb89e8822c12f6355e70286e7254c70994c75cac", + "source": "aya-original-human", + "source_text_sha256": "00423f3010c8b4a2bce869240ccd5b53d192ee404e87365988580033c3e7e975", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-3", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "00b7c33b5c33fcf81a988d62aee83b5e5694e2ea2a50b48bac62e408572d184c", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-4", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "011c2e6ca12666586b1c6a0c8f8a6176ba917bc84e26834ca8417c8e085bbbc5", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-5", + "language": "zho", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "source": "aya-original-human", + "source_text_sha256": "01c30bf0715675af43a0afc38fc8db0e0d261973ac6744efbb27a8025295898b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-6", + "language": "zho", + "group_id": "aya-author-68f2ab39801cc0fce2d39ceedb89e8822c12f6355e70286e7254c70994c75cac", + "source": "aya-original-human", + "source_text_sha256": "01e88274ef462f2a9d01e137319ca41e44f32f05386304692d55e394524f45d7", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-7", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0201a4007ec50f95bbfe24a56d091805554b962fc3028c7892ff382f4b64f403", + "task_labels": { + "domain": "math", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-8", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0232050e0c9b09bdf02c3847591b3a1feef3ad17dbbb5aedd3ee3fef8188b05b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-9", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "024e839f4178865eb492d8b89cdb56f09bbe6edab798245876ebd52bca20cf09", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-10", + "language": "zho", + "group_id": "aya-author-68f2ab39801cc0fce2d39ceedb89e8822c12f6355e70286e7254c70994c75cac", + "source": "aya-original-human", + "source_text_sha256": "026e066375e20e135f45e3a3900c4b3bd95bdcdb6b7602c67ed9cb9700ee5677", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-11", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "02afd6f1b83edd9a519b7e37fdc2bdacfcdd9316fdfd093adbe7bae375ba5ba4", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-12", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "02e40407b68eca66eb764e6c60ba496b6522a2e0e4f7fa2da196341906cbcf53", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-13", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "02f25b2c9f16993c3504529851276992e4f462a55bffd1845e43ba90c0b5edaa", + "task_labels": { + "domain": "chemistry", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-14", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0300a143f8df321cc7635b8aa9867fe04cf48e82e04520e52a9819dac6a30d5f", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-15", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "030440d17e910ca70f72100a0eebbebc767c25a8f7b580108f1ac7ae1fcc8080", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-16", + "language": "zho", + "group_id": "aya-author-a1f5c118a90749b3f3c4271944a1055c676aa02cc0f6bdd26e4cbe986eb17a51", + "source": "aya-original-human", + "source_text_sha256": "0319ea6c57add9338b4057358d17591d6255c99b9e05366b99518d9df768abc4", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-17", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0321ea3deac6561040e17dd623df2740e75d5bcf62fa04ea7d8a6593dcb4d2cf", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-18", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "032652c3807073c42131f8541a0b6718fc9d9de26e056c33e71ef5535f93d746", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-19", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "03534d396ca63ff5f1ac496baafdb4568e2d86367000c2b4ea35aa900e1c94be", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-20", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0368405f0e86bdf7acb14776e6412f2d65054d58c9be2dbb26ba5005ca369476", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-21", + "language": "zho", + "group_id": "aya-author-5835c5271f7c7dce1ce63632094653841194bacea7b48630424f44ddedaf620b", + "source": "aya-original-human", + "source_text_sha256": "037ccfbe94f8e5451b7e80d666c318651d35e48af49932b7d7db6f10ec9c4117", + "task_labels": { + "domain": null, + "factcheck": null, + "modality": null + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-zho-22", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "038e30753c7de1c7ad1473da0bf022224e0a629b13bff78c70ff2615f2b183a1", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-23", + "language": "zho", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "source": "aya-original-human", + "source_text_sha256": "039ab4e775164001ba1b3ffb05c4a46cf00733517d1cb6ebfb201b6419f14d3b", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-24", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "03b65b7b52fb00c90c1b7275c2caa0bfd3db09e67ad455402ec7844a14c179ee", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-25", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0404c1fd7459817def485e8d3e86a23681ccf1fe6332c7882c15022bf9c18758", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-26", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "044f5fbde9417a1e03d0f81e4342d1c4f388218fe4a22596339d57d626ac1410", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-27", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0450e9f802c5a6bceae305a83499eca2f78dbe02e2b8e8501f7e7db8385c3944", + "task_labels": { + "domain": "psychology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-28", + "language": "zho", + "group_id": "aya-author-68f2ab39801cc0fce2d39ceedb89e8822c12f6355e70286e7254c70994c75cac", + "source": "aya-original-human", + "source_text_sha256": "04738dda49002ad731e5dd94c8f52ca05ed9405077642c28b7cf270b0482c7c5", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-29", + "language": "zho", + "group_id": "aya-author-fbd65cb390a74fa13ef2d0518e464c83e90fd1df5a5b2da91d8850177ee1b3c3", + "source": "aya-original-human", + "source_text_sha256": "04b4e860af77017cdb646c7ab6bb52f7f70ab0babb0a752b52f29da435f1ca67", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-30", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "04b7c6324f86386b8dab1e93bd54cbf64f76a85e54f383fabe72d5d102ee1345", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-31", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0517dd4626f178e7850a868c600de2f738b040c0bd45f0717ff57f8d02fb6f25", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-32", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "0583a2e6ae573236f3cf71fb6acbb039be46e8da7bc0def3952c52f8ea6dee8a", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-33", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "05906ce297a4be42cd969620a2269dedd6534db72b33beebfa1764fb63ed25b2", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-34", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "05e5e8988012d73321f44f4ba664022930d50709353869b6a7a18d0b589ed70d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-zho-35", + "language": "zho", + "group_id": "aya-author-e815427bc2b34d1ac8784d85a370cee46e094e88b21363a272494690d07d2e22", + "source": "aya-original-human", + "source_text_sha256": "068164dce9f105d210b840cdabf47f40f8e4322ee4bece15ca0a1ddbf92454e5", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-0", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "00247a3d5eaf584ff4d3b249e3b596669a4b50b775179e5cffdcc2748ea469e7", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-1", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "0059dbc49c4348272f09ff359306848c19dc0953c6c1b64f40555e90ff3011a8", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-2", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "01284f7b0b71f55a9db4dfdd597249c2dc999d78a5c020f8f865bf6ace34f60b", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-3", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "01ca9573b0f832f51c4a9612740170ad456ec77e650385f8d889c74e37fcc392", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-4", + "language": "spa", + "group_id": "aya-author-8412f807572f18b56680969c87fa2f9a99fe2f1465df00350abca570704c0cec", + "source": "aya-original-human", + "source_text_sha256": "01d5ebe80040d13604b9429faf711d13d9c642601cf89450187b2bdc650a8dbe", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-5", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "0207509cff83d6a681f127f7bf79dd94f62f5015f64376ea4db59f3213ef3d2f", + "task_labels": { + "domain": "engineering", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-6", + "language": "spa", + "group_id": "aya-author-5ff707657b272c932a8eac9bdfc5d359da9fe3698e3cd1e6125df4bfa6bc331c", + "source": "aya-original-human", + "source_text_sha256": "022eb82ad31b2db5bf53a861e75190b1f6c5e51cb662e21d816726a31f7bce89", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-7", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "02633527a343ced35736e72be7e598f1c7e4c0b3c05cdf8afa2e3fe986a53dca", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-8", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "02cac547a2b089c5a2650b8ee6c4543b3bd0ae18315d54975729ac42cba14f7d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-9", + "language": "spa", + "group_id": "aya-author-128fbd25726b0fd15da30c91774f2c58f99e408721aa4743579d9eb7f91a40d0", + "source": "aya-original-human", + "source_text_sha256": "0328e278611b7affd42135f55e027671903d5d46d59523d95f531f7479c5d862", + "task_labels": { + "domain": "math", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-10", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "041506e3749c6df054f67bd285dc82966daf550662188363f94a5d96d6994735", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-11", + "language": "spa", + "group_id": "aya-author-8412f807572f18b56680969c87fa2f9a99fe2f1465df00350abca570704c0cec", + "source": "aya-original-human", + "source_text_sha256": "04926234bde514bfde3b093f63b5be86c4f0513acfee1597473ba13eaac804ac", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-12", + "language": "spa", + "group_id": "aya-author-4f06d092131f78ec23fb57c324079a0adf08967b95883e68356ecca0bb75a97e", + "source": "aya-original-human", + "source_text_sha256": "04b1d8d7dfa032f82869ba7d4bc32dc509854b1cfd98b18c26fd3f852d272878", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-13", + "language": "spa", + "group_id": "aya-author-5ff707657b272c932a8eac9bdfc5d359da9fe3698e3cd1e6125df4bfa6bc331c", + "source": "aya-original-human", + "source_text_sha256": "055c53390ce747230fd57bb4b37dac0a5b2b0206048ce6bbe26c64f24a5b9647", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-14", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "05c79fb60008d1d9ea8adc627f395fee2231cc578e6f5867807feb835647f144", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-15", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "061b79a13c1a20c9d513a92c169fc09f2d72960dc7157225283ce16cdb09adda", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-16", + "language": "spa", + "group_id": "aya-author-7122484da0e5ac94986ec31d4218642de7a8510a84e3b773515061de6d8ffa88", + "source": "aya-original-human", + "source_text_sha256": "0636f16d68a0955ecba91d21d2e5139e27e7eb19c0141d1af35f047ab32eaf8a", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-17", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "065c667039a3f4f294aec575e2621a1c52b5908c3dee3a65a90a04dfa3c34b27", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-18", + "language": "spa", + "group_id": "aya-author-2986e8162ec136cdcfdd5402e3765f8fa028d4f1726dfec5e501a6c650f5adbf", + "source": "aya-original-human", + "source_text_sha256": "0677b4aa5e7a59336e9bbc767d8d8f505f05484fd57e3dbeb24eb1c6fe38fc49", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-19", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "06e1ed65e86923648bd3bf337b63e1823eb15a80f7b80cbe542113ab16aaadb5", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-20", + "language": "spa", + "group_id": "aya-author-7122484da0e5ac94986ec31d4218642de7a8510a84e3b773515061de6d8ffa88", + "source": "aya-original-human", + "source_text_sha256": "08015c9411b2fc00fac054a61e4c15b43185fd97169925c660948b43703639bf", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-21", + "language": "spa", + "group_id": "aya-author-a8dd08a8f2f5850dc68fc2aba9e3cfe6a6e27a81e50136717c60d7f003641cf5", + "source": "aya-original-human", + "source_text_sha256": "08a7f24e0f33ab84887c657f1553673d1a97750ce3c982ffe7dfd13b11dd94d8", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-22", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "08e3818cb3dc42e77cd2429af8338a387c301c26da096edae7e24f0718a60711", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-23", + "language": "spa", + "group_id": "aya-author-5ff707657b272c932a8eac9bdfc5d359da9fe3698e3cd1e6125df4bfa6bc331c", + "source": "aya-original-human", + "source_text_sha256": "091f11b10d928a5870fbc10bf10c1d6b299366c75d3e649063cf82c9f012f463", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-24", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "09951709ca65c5834cf4340697c003ff41156bbe631922e5e232961be6afc3ae", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-25", + "language": "spa", + "group_id": "aya-author-f648edbe7cbbe231244c420e627f97d92f71186e03cabfd628f589afccb168a4", + "source": "aya-original-human", + "source_text_sha256": "099d23a793c85d35565c464a1437e6adbe6658bbe290bc781a729ba291d1f0f0", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-spa-26", + "language": "spa", + "group_id": "aya-author-8b952b20ac15bab5342d070620109cbbcd07edb78f7c7ae9b655edfc8f3c1f53", + "source": "aya-original-human", + "source_text_sha256": "0aae8aea475af35808451de1f75a0237e0e66685e3d76905d6ff293671c4085e", + "task_labels": { + "domain": "chemistry", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-27", + "language": "spa", + "group_id": "aya-author-eff3c3f1ba2cb27d2bdf365f30a2c3d3686f677a896c1e01ba7232bcae23fbf9", + "source": "aya-original-human", + "source_text_sha256": "0afb12dce7c1572af56f2703466cb06bd09975015d96dea290d875ebba8f363c", + "task_labels": { + "domain": "philosophy", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-spa-28", + "language": "spa", + "group_id": "aya-author-a8dd08a8f2f5850dc68fc2aba9e3cfe6a6e27a81e50136717c60d7f003641cf5", + "source": "aya-original-human", + "source_text_sha256": "0b2a71d8676a08dff76b37a167213c5b47518b3b306f58aca8ba0b217338bcd1", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-29", + "language": "spa", + "group_id": "aya-author-b5a64acb53d11eb8ff670da1ccc9b5e0dbc4194076c747b3e8e223cf3aaa62fe", + "source": "aya-original-human", + "source_text_sha256": "0b77394fd2f96566efd7fd8de6afb61678b1e551cef3c48d8fefa9d61a08bab7", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-30", + "language": "spa", + "group_id": "aya-author-4f06d092131f78ec23fb57c324079a0adf08967b95883e68356ecca0bb75a97e", + "source": "aya-original-human", + "source_text_sha256": "0b81f8ee86679b80236a69048117aa59ddf0c4b2ac67bbb21cf9613f0cf2c21e", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-31", + "language": "spa", + "group_id": "aya-author-7122484da0e5ac94986ec31d4218642de7a8510a84e3b773515061de6d8ffa88", + "source": "aya-original-human", + "source_text_sha256": "0b8278687899deb178f20ae57f572dac70b877d83a8aa2b93e14f6331d864dff", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-32", + "language": "spa", + "group_id": "aya-author-9c34ccb2b0cf1f0411b88aeffdecdf4836f6acc5111e86bd272f1f6614098181", + "source": "aya-original-human", + "source_text_sha256": "0c3ee6761f0f49573fd1d2514229862a9a995ab6a0b4b97a7ae8a1a9659e855b", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-33", + "language": "spa", + "group_id": "aya-author-8b952b20ac15bab5342d070620109cbbcd07edb78f7c7ae9b655edfc8f3c1f53", + "source": "aya-original-human", + "source_text_sha256": "0c3f3eac70ec5ce31c9024337cc9fe5aefe82e14a5eea3e96e44737426aa0e7c", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-34", + "language": "spa", + "group_id": "aya-author-a8dd08a8f2f5850dc68fc2aba9e3cfe6a6e27a81e50136717c60d7f003641cf5", + "source": "aya-original-human", + "source_text_sha256": "0c7626d61458f786975fbfcf95c0aa1fd4767de00b7ef33168684bc4af7b9cdb", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-spa-35", + "language": "spa", + "group_id": "aya-author-8b952b20ac15bab5342d070620109cbbcd07edb78f7c7ae9b655edfc8f3c1f53", + "source": "aya-original-human", + "source_text_sha256": "0d1b9531210587d19671b371e46c811cb3140cb9f81f5f5a56c877f54ff6c86b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-0", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "00085a70e04ad5f9619e6adc9212172e9a79f960898a897b7081c817fb41dc65", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-1", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "0025ac840f4c3dcbeaa782176a170e4b5da9f09c051bed06df01515442fa24a8", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-2", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "003400e3ddc0deae7888904482ad57f739923e4e3143c6fcb6047d940bdeaf6e", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-3", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "0052ec8b7394f6f4a8bfffb120d920a13c8a135cc072439d210385c7494a471c", + "task_labels": { + "domain": "chemistry", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-4", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "00c15d3d2f8467c6d89a18656cd9413272bb5ab08b4ccf8ca396ccf91d359743", + "task_labels": { + "domain": "chemistry", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-5", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "00f9adec6b3ac9eebd05015037fd17b521104c9275b4870f3b3564ec5f642933", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-6", + "language": "fra", + "group_id": "aya-author-1245db41ea2e3a92042cbeb15a6dd7230ab562db06b9d6286e7342fd442e5fd1", + "source": "aya-original-human", + "source_text_sha256": "0148759121374fb00b49a8ee304f3f27bdf79435f9cc45406cd2608997908f12", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-7", + "language": "fra", + "group_id": "aya-author-a7d34b89888294e1afb42fb9160bd645ccfa37b7a1ef8701244bdb1605e572b6", + "source": "aya-original-human", + "source_text_sha256": "01dc955510244764c3ea5a5f2b26a1bcf63631a8e91dd5dccf38c9727a1c2651", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-8", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "024cea4ee9b3b7b239241552f0cb613cf21bdc499fa3f50f5e1341a677559d56", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-9", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "0312504d169c0bf9b05ee14b20153730dd96745c2a756d93cb4792c3f97e35a9", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-10", + "language": "fra", + "group_id": "aya-author-a7d34b89888294e1afb42fb9160bd645ccfa37b7a1ef8701244bdb1605e572b6", + "source": "aya-original-human", + "source_text_sha256": "032fc2ca475912edb28875a2911e945ffbd0957da43ba119353e34a5fc89d3c0", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-11", + "language": "fra", + "group_id": "aya-author-a7d34b89888294e1afb42fb9160bd645ccfa37b7a1ef8701244bdb1605e572b6", + "source": "aya-original-human", + "source_text_sha256": "0368bc70c5811a90e48c742564a3ddff6151e1fb0cc1d9cb0cdc707754afb37a", + "task_labels": { + "domain": "engineering", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-12", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "0434346f2f814767fcfe7a876813a9fb5049c68f1a6f55eca2463ecb78f15e5d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-13", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "04822a5cec14a13b56e651c8418f5c842a1b0dbb757918077ae9b311f92c21da", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-14", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "04a8abcbaf2fbaa0bfd91f53b58d31c855971019ef223d72a8cb001ef3863cd3", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-15", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "04cbd1ce822bd71caf957abca31163d39a66f7ceefe7c688ee8dd163eeaf528a", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-16", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "05205a83cdb9b8e5cc5e621929353de70417deff8a5f73c500e2b5d206854a3d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-17", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "05b1e7c13a66897d5aef6c595996ea4f1467ca744051394d19528c3acee7b918", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-18", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "05f4a3a3448c4da53d764777ad1d44d0168819bb66797776d27e2fa260e0966b", + "task_labels": { + "domain": "physics", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-19", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "065d519832f1fefbc290aa69d4c38090c191c78ddcf5f7b17fc8eecacef4db6a", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-20", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "065e0f5478202978e2a1562e7ea71b33496e09fc4b4127e82b7b808b4b4371f2", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-21", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "068bd805ec24cc6566029dc28b04e768625bb35d06fbb691aceb8c1bada4790a", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-fra-22", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "0697f692426b6fb77c274185b060b454cbce83c4fd406c9356249465fe2fc1c9", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-23", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "06eaf7d8ae23ab4457660d109c9ad6df77e957f7fd2f34b1cb9a6367ade916da", + "task_labels": { + "domain": "physics", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-24", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "070fb3d1b82428fa7cda9022f7454992ab5b767d85f30b7dee6863835cdecab3", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-25", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "0713779fa0514fcd6ab452fd8756c64685b7079450d0b7093fb130f36deaa87b", + "task_labels": { + "domain": "physics", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-26", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "07625ef864193d8a8765c87fae1cbadc8ce7531cd64c20567a2829e76e8328c7", + "task_labels": { + "domain": "business", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-27", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "07e067798e2193f8a6f9d65838930fa52a606da4ab17fa0025b484c887a439bc", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-fra-28", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "07f14c9231aed94a35d25bf7c34f6b8d9eeb0a715fbe1a7bc416533a8503f01d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-29", + "language": "fra", + "group_id": "aya-author-54d94d4a442a2bd7562c58ec5ffd3053f6d2a82b17f0315354e75bde8c03329b", + "source": "aya-original-human", + "source_text_sha256": "084908bf67ca11d7ec92b5515f3da4b74a9ad3ed3bcfc5fbb7884ca53697b8be", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-30", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "08a396e0a3b9d6d25499554e4db7426471f9c4e912f12d0ffe05decc5cab3b31", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-31", + "language": "fra", + "group_id": "aya-author-1245db41ea2e3a92042cbeb15a6dd7230ab562db06b9d6286e7342fd442e5fd1", + "source": "aya-original-human", + "source_text_sha256": "08a92893ee6537ed8f55161328cae4af44f9916f3642bf8782e07673684d8f34", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-32", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "08fd831d21c2eba14d1004773d2c57d757a18cc4a5e981a7e4aa89cab4e7ca6e", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-33", + "language": "fra", + "group_id": "aya-author-30c761354de060343b18563795c46d572ef45e51b28f2977f3554b3815fce67f", + "source": "aya-original-human", + "source_text_sha256": "0917568584175cd9da38fcc9755b1d1d1014d41bfe956793cfb41dc40b2972b6", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-34", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "091e35e53f910139fc15cd1c0e5fd734aefc2a609466fce8139a1e911a73686f", + "task_labels": { + "domain": "physics", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-fra-35", + "language": "fra", + "group_id": "aya-author-7b7981c86fccc816e2d37e35ccdbf3141ea403d7281e3381b97ae61ad1551e27", + "source": "aya-original-human", + "source_text_sha256": "093e4f8b1251c77b303a9e6a813f3596adf87ab440b1ec0632aa1058160c2aa3", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-0", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "043888460fb33d0b759c0e175bb07b1a929f3558818124afa5c8bbc758db8e0b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-1", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "043c8ed0be92b01a5cc4ff5a61972749cd12efe9bb700ce65dc5119ee0ecffe2", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-2", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "04dbd881c93195e8e6f35dbc97e4abd79e3af1e7486fa06b876da343d3899b13", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-3", + "language": "deu", + "group_id": "aya-author-1b6b66d80b6b38f2b24e5bc045480cbc476ff0faa7fccd210f012a58360dafe0", + "source": "aya-original-human", + "source_text_sha256": "07fb61786b605afd06a8c7c377f9dbd33893f3405232d527bc6fdd61d58765c9", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-4", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "0b50558884acf7a26ba5ab102f87d3645800cd66870162c611ad4e2164ba5a34", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-5", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "0cff628fa984ff4b52ad8f743d9f1666f7972989a84f1f1c8f7ba14d64d67bde", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-6", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "0fe2954263a5a01c786acf1f250fd45c99d841b2ffc2a7be29d04cae15352de1", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-7", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "15aa59d101f164584875af0cbf18bd216d250db5375734872f16f6eec4d0b2fd", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-8", + "language": "deu", + "group_id": "aya-author-08cf632b35df86fe1b539fa75107c5c6c9a73f4a7a912cf5c97bc0a75838e211", + "source": "aya-original-human", + "source_text_sha256": "174c7135f7c96ec09547231c9479bd0e1a5b174b8bf30a3a3a8de847ac104b69", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-9", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "191d0ef78161750e679291fef533fc575015ad56da26023a944fd3c5481229c5", + "task_labels": { + "domain": "engineering", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-10", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "199388a60f9ca410c39e3fdea97237af6af110064eb501ab006477745fd35696", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-11", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "1a7cebff09eb27956b45256dbf190f1a99fcd3e345fed97e44292d479495a4ea", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-12", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "214dcc0773dad1c0cbcec25f24ba65bf749ee2ea1ce7cb2a1519ac38a90e4996", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-13", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "21efedebc2cdec42e1e2ba00e78647817b3a366f225a97fcfaf43960786d7577", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-14", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "2389035101e4b69e6de00503fcc3ee1d1a6398791c5564383606a4ad0d90b308", + "task_labels": { + "domain": "math", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-15", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "295fcd3201436f08177510fa76b681d062ede5707641b74ed2d57213e8eecce1", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-16", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "2dbae648f3da6bbabe499d647fd0a5c1ed0820c7450c8e1796ac25624a3c62fb", + "task_labels": { + "domain": "engineering", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-17", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "2e732ff841d45b5dea4a4f2bc0514e0a736a8fc2d5000cdf562c233d86880aa6", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-18", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "2f0338124237227f96ff10a80cbea874e6e4685e5b93a7b640a5ea327b67eac3", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-19", + "language": "deu", + "group_id": "aya-author-027809a6144dfe3ea9da473f1d2b99e213697683ef07e3d4fc005b433da2b28b", + "source": "aya-original-human", + "source_text_sha256": "2f6fd43cee6490c64023b0dc95f25b7171d754065bb88b30a70fafa34b38c230", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-20", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "38ed2570b2e4cf852a57219b34d96ec7fb408547c68ae46c88a8c5cbb85ee3e6", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-21", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "3953f89511c610da41ebf66a801a1a1986bbf1b531a0aa5597fa7f5dd6a5eb6e", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-22", + "language": "deu", + "group_id": "aya-author-9a919ec5f7307467147ec593992cdacc0dfecd24b1fa6d09f75fd81675fe29a3", + "source": "aya-original-human", + "source_text_sha256": "3f8453cf70e9156943ad6795a50a14ad54d9c499295fa83fe3e16a3baea30245", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-23", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "4188b79265121ea5efe2f3009575a1223276792235203d72f84c7520d2467857", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-24", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "43e9e52cddda520b696cdd23d561d9c82793898ee3f7a400d47b047fc23a3a05", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-25", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "453e94d980859dea5e0088c367322a3b307ec2f4ccafd74701771fd4417fc64d", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-26", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "46594dad7d3524f13a516a02ea6c628b0c0cadd35aa18e4c773c7919ca7de329", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-27", + "language": "deu", + "group_id": "aya-author-89da884608fffca31bdfae1e82eef0f2159a3cf0c47f6d6ffe01f25c6aa6bb42", + "source": "aya-original-human", + "source_text_sha256": "48d327dbe70a57d2197708570d68cff3d6f8b6d09c400b4908be28a4e7599f59", + "task_labels": { + "domain": "chemistry", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-28", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "48d8a29b276a0b192b45cd45c457d24f0fc8c7e8b4be34ca8cfcd0c3402b19d8", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-29", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "4c8d12ee094891bc679a06d7e56ef005c32378d39f463ce5acb3de562948ae2f", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-30", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "4f30e58caa214de86cf10fa75d3f75cd51043f1ae73f509c2d01f7169983eea1", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-31", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "4ff40966ca0e02dae9d3777d0e00f39d2a5a25980aa477c7011ed3747ec656cb", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-32", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "52b295a9df015e845fa1f19bda79f5a8e7eef8560b8e8fc6defe3482fe1b0599", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-33", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "56a24ebb548950b623142e821968ae3804098d6367663627062d264b3f2ca445", + "task_labels": { + "domain": "health", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-deu-34", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "581053636fb46ec87c84feeb26a1fef90f0287c9383d0ee1f5e198b7dbd8662f", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-deu-35", + "language": "deu", + "group_id": "aya-author-58d762b3fa4827f7939a9f35504321e6d2ff337e6abcf3f70dfef1115d088b5c", + "source": "aya-original-human", + "source_text_sha256": "6248d14df6032a1071e7549163645db5d61fe07b8908dd6bc330fb92b6043888", + "task_labels": { + "domain": "computer science", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-0", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "000415d66c2173bd5686c4ff4b09a4c575fb0835a91e38d16964e16b44e272a8", + "task_labels": { + "domain": "law", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-1", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0007f85a418fe1b3b3f3bf0bc4044c345b8bc88134b0a966d1883f60f1356468", + "task_labels": { + "domain": "other", + "factcheck": null, + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests", + "exclusion_reason": "Incomplete request or verification need depends on unstated response policy" + }, + { + "id": "aya-review-jpn-2", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0013b430072caf73eb9345b28645dc845ae2a73b33cc30f43ce9da89b4d14356", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-3", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0018247612dad4230ff7daee24cf9bac5b74ed0cdd7256d3ae1eb2fdaa2f91ba", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-4", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0019dfdb104e7f290cdee5303bc417d96aef0e02f74bbb96ce5847425683795b", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-5", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "001ded3a9bd75fe78de4146d4eb17a3b5a8a11e527be43527c3173c9b0c9c6ea", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-6", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00237acfeaa21ccfbe8a698c005289a7e369213163fd880e553ad73fdddb33de", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-7", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00299121f2e206d0c3e75e2e833b1bffbe47a51c798ad34a13019b7faf3c2978", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-8", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "003580fae6e2ce02f070ca527c40ac8b2d9ba1232e91b811fd4c6b337dd728db", + "task_labels": { + "domain": "health", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-9", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00394cb5ae00eadf12803c996afc3f7c35dd1eeba6a9d0ee55a7ed2887e68d0c", + "task_labels": { + "domain": "health", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-10", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00573c03b3bbed9b43ac9e611abc7504ef347add319eb32d5c155990617bd033", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-11", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0059cefaf5b2ab08d0ca74bc5791741d6cb51611ca52b2697b21d2b391b987b7", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-12", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0063dddf4f617a9ed0c1fa74e9af1ec142cff426288eb84081d467f14ab9f3d7", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-13", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "006527af848ee9e6d91dc2714d09eb0b562c063382484721b4615b6a7f1c0c00", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-14", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "007bb297d185d967e7175621a73821f591b930ad0f6454ee41063fdc378e0d57", + "task_labels": { + "domain": "health", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-15", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "008a4f5276c0fb05e4a1668e10fb1530213f672db8dd6ef5dbe1cf743fc66e40", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-16", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "008ac923aacc98b1f6dca0d58e4c430c5e79c77d6c39eab61e9bd00bf48c1cb5", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-17", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00a716f8dcf9bd0242b79de5d3d7291d34a0651c321cd6a6d215b0f707188882", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-18", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00a7bb8f8e305650198716af6f06909a68dc2c5b3c9bc9d50fb350d5653fe0af", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-19", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00a9c8548a9b215d8d38308b5b042f71a4b786297d510ed2b696a59c78904beb", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-20", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00aabd7b43350a561a2fe8b5fcf033b55c1c9ec8dae9ab710351a2fd61cb18a6", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-21", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00c7506fb5beea02ac784c015c50849bbd0342e66e69c2a5d13f02457414427b", + "task_labels": { + "domain": "biology", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-22", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00d58181d84832d459bcd21a93ab50c1219f00e0c5bbc1769b02a3875e98239c", + "task_labels": { + "domain": "business", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-23", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00d77fa18658b52f3c693d216efa2a360cd456559e7f4387aef4567ee0af1152", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-24", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00e30a35e19ab05748c22b89c43a7e080eefc62957faa480c47080e67a4b5991", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-25", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00f1f76abb07f3168c5b29bb6878d5ab29a2550843bd57fab7d97b3c167751b4", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-26", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "00fca2b18255c1166e7ec779cde18caca7759c8a78f8efc53856adf9f65ad73c", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-27", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "01024869b8a8eda5bacd1842f11fb442072f89ead491f35cb45e9b2c53f02569", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-28", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0116a27888882807436488319d9504f56519a26e96735a1d481f1f2cb692811e", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-29", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0118e914113a9f45ee77b4f345f72420e91231801796346c680ade8a2ba2db83", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-30", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "011abd13f75fba618468116000ec17e7bbfcb218262b116a76aebbba42958bd5", + "task_labels": { + "domain": "other", + "factcheck": "NO_FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-31", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0121a0613ca3c1f6ca59ca01e596845998552705c2cdf3e07ef6b9ddd50250bf", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-32", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "012528005ed1283f39a05715fb1897ebf91cae95a32bbb203211a03a913f5222", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-33", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "012bd537551bcc54732fddd0e54c0e3c1c8b97956d6036a7be5edc308b9d8013", + "task_labels": { + "domain": "history", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-34", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0132ec1747d5f024a1e69ae5974ac43d100552afa3300eb86953613df5b0cfc1", + "task_labels": { + "domain": "business", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + }, + { + "id": "aya-review-jpn-35", + "language": "jpn", + "group_id": "aya-author-a941c8b29c9a082c6dd06d0e49e121e1d522d826868cf6d8241681e17b417561", + "source": "aya-original-human", + "source_text_sha256": "0139b3595119321e6be541cda3bda682dc447c39e8191cf46d6b8f38bf56098b", + "task_labels": { + "domain": "other", + "factcheck": "FACT_CHECK_NEEDED", + "modality": "AR" + }, + "annotation_provenance": "Assistant-reviewed task labels, before model predictions; original Aya human-authored requests" + } + ] +} diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/dolly-fact-reviewed-v2.json b/reproducibility/src/training/model_classifier/sequence_repair/data/dolly-fact-reviewed-v2.json new file mode 100644 index 0000000000000000000000000000000000000000..5ca8dfe5cebf4d9d270581bd5b78dab1da051114 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/dolly-fact-reviewed-v2.json @@ -0,0 +1,2908 @@ +{ + "source": { + "repository": "databricks/databricks-dolly-15k", + "revision": "bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a", + "filename": "databricks-dolly-15k.jsonl", + "file_sha256": "2df9083338b4abd6bceb5635764dab5d833b393b55759dffb0959b6fcbf794ec", + "license": "CC-BY-SA-3.0", + "attribution": "Copyright 2023 Databricks, Inc.; Wikipedia editors and contributors for source passages" + }, + "annotation_scope": "Individually assistant-reviewed task labels on human-authored source requests, assigned before model predictions. These are not human-expert task annotations. Excluded ambiguous requests are retained with reasons.", + "records": [ + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9b3499032194045ead5c64e710055b019a91a4272f6f301ddcdb89ebd45a6a1f", + "id": "dolly-fact-review-5311", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 5311, + "source_text_sha256": "3b9703bbffea50bc245fb3e28eed3e5d8e67d97ff2f56f5f21a5d674e9390234", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ecfd82086193e8a62049b5c99f170da2075edae1e7edc403148ec283fd5f23ca", + "id": "dolly-fact-review-7248", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 7248, + "source_text_sha256": "b10615a576b5b0d90b59153ee3e6ffd6874b90f0ee43624766a29ba412a5ded9", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0721239b11c5419d3ce0d594e8f73ec9da17b24c8a4237ccc1de5a967d1c963d", + "id": "dolly-fact-review-10270", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 10270, + "source_text_sha256": "382d689a60eb46d0fdb79816fca9688571c5d6e6c8060847dafa9e14a158d1cb", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8751be9fa09d057479b18a3c873855bff60c87861f519140fb00da2242ee9e67", + "id": "dolly-fact-review-3520", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 3520, + "source_text_sha256": "a1568576c3d79fddb6f69e95ba777e08a74bb3e396b3b90916bb0d0fa30e6fe3", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1462f078d1d23b27a4f9722a81a730365f20e600a04fad971882522adf2a632f", + "id": "dolly-fact-review-4322", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 4322, + "source_text_sha256": "10adee5914d63382959fd342f78d38659ba3354988703a3abc369c0eacb6e16b", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-03cbbc9ac6d78ec9010779a8b26b2e4769a5c04ada35a9c0b076dc8156b530b4", + "id": "dolly-fact-review-1949", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 1949, + "source_text_sha256": "7a40249c404286d7cc9586654b657e78eb32b5aaece3030ae4da7b3cf04a6259", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Heavy/light threshold not specified", + "group_id": "dolly-source-626b441123dd9df14edf1e539ffdce7d51b7fbaa304e6547a61245d7eab0123d", + "id": "dolly-fact-review-552", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 552, + "source_text_sha256": "40f14ae8e24906e9c7e6ae4f83743f0165a98f430732b77c11157089b72f9696", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-da976685f74a28c127941562ce8dc566cdbd49f16cbf7c9deedff6497f6260b7", + "id": "dolly-fact-review-1978", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 1978, + "source_text_sha256": "d47976a6868e67346777c93ba83f99e9ce1ecebd52107d4279f905182fc78994", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9ab99949313d70fa51016e0794a874b9ef26e3e42a3bcc4e27200c5c5bd564f3", + "id": "dolly-fact-review-5018", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 5018, + "source_text_sha256": "da7e60bc7bd85673b122bb460faa2161f4ae0b5c9f3567a22e959c4ed4bb93c8", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-6fda296623c5017806ffd5b15c9517af5e0b226e142fe9d49454ce497c34d77f", + "id": "dolly-fact-review-11581", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 11581, + "source_text_sha256": "66c46dde9ddd7b4fa9b07d8fffef34ca238f7b129f2a167110a0a074504b151f", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-4d51e5a7b806ab663b82ef3b6e374e8c2a5660b10559116bc0f19bc65fd69274", + "id": "dolly-fact-review-1461", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 1461, + "source_text_sha256": "0c12a196e9db6f697e7b415ca7dafb65f081b5026a34173367e521bb66c8b796", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8337000cc3e8a5d97c7e193fd64fd83fe8798f0639db264decb926a8111e725d", + "id": "dolly-fact-review-5717", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 5717, + "source_text_sha256": "60126eb4ab4cd87c7a5e64e3dfeb3d2d2875744317dd5a190d2a108f5c8421a8", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-5c2fb1ba6a8e8fbece55dd298d99dc638ace0a60d41c1e293aef782102e2209e", + "id": "dolly-fact-review-11478", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 11478, + "source_text_sha256": "cf9be53c60a53a1082dacfe4f1f9233f35d3c64cb8dc317c43e2d329b5158fd5", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-5a731f0942cc10c467d8fb3f75e26e5f3a9f78a95e832c61adb95ef27a37af10", + "id": "dolly-fact-review-9951", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 9951, + "source_text_sha256": "80b92de07c3241dc1df4303765b62eca8635c1e69c8bc3385f637e6d7588ae30", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-57b9977d196aee9b5a09cbb1c5fd6cee38686a2805a55aa8cb1a9134d8509ad4", + "id": "dolly-fact-review-12392", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 12392, + "source_text_sha256": "43c698d96e374056e33179ad62acb481686894e43cacb89d0de71b71450077ba", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8aafd8edbd7622f00299963b05c62e8ce3f201395fb96ccab08ae152804b3287", + "id": "dolly-fact-review-1342", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 1342, + "source_text_sha256": "a6e8d76058d5cc5106a8300919785c495cf2e3dd02d217aba9ef32ce081115af", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-c8c0fe5be98e74d356a6a61ab700648fbae4a8363c6e1f2bb03870e66003d63e", + "id": "dolly-fact-review-14115", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 14115, + "source_text_sha256": "91e01e8e408057ce5b206d0b6d32748aac5d3f7a1fef4f537192f9552c8b4f9b", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2d022cd1d6c713d9e070c7ed09db6f379e54a1e6b45c8b2c2aaf345455acd9ec", + "id": "dolly-fact-review-9797", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 9797, + "source_text_sha256": "8e6e66251e271ff3334bf17741140dc3c57b8bb5ed88fa437b15a22b30be3cbd", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-aed7906abe8fd31b1b828fbb8cb0ceedf770176048221972b08ed3224c2e68bc", + "id": "dolly-fact-review-2791", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 2791, + "source_text_sha256": "8d9a4297ea9c86eff504ebee5ef90bd2e1853605536868421f2c150d311d107b", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-498706f22adfe03201f177af5ee762bba0733f523de54c31a9445458ecd2f526", + "id": "dolly-fact-review-3004", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 3004, + "source_text_sha256": "d3840f57317bce5fee53202d26a5e991a6a819e0b675edff8f4096c17b424b2e", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9c98d383a7696e814ad5b0a7f85a8141acf1ba06c3e2399c9230ac2a8df2129d", + "id": "dolly-fact-review-11193", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 11193, + "source_text_sha256": "473d62042fc13d88743578acb37980bd4178d1facbbaeb2d9923abe53ca52a37", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-27f2577f6d1714a2bede4c1c9fa187101ab6c5e41ea13bd96dcd86170c166dc2", + "id": "dolly-fact-review-14713", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 14713, + "source_text_sha256": "b013ae02040e37cc912fcccd1a29d3738e4e6648ce7e4183794c9544d48d0a5c", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0c2b42aeea9e4bed918aca53fcaa6f823727eaccb125e6d6cd1d6965bf4c2059", + "id": "dolly-fact-review-5460", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 5460, + "source_text_sha256": "2327fc9bd9b6eece44c17c2b707ab402a3ea2b67338fbd6834a1767ac124e5c5", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-81d94f3a419e6b418d507a97cfd2004788733da7c459dec4eba66dbac55b5acd", + "id": "dolly-fact-review-205", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 205, + "source_text_sha256": "f425e5911f23a44554606270ff4c99b0d686eaf271e987f6bba6bb72ba876db3", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1ee0212e35c6825172c4d3a6524ae2959a98b55edd6c24214f9bc9e40c88268d", + "id": "dolly-fact-review-556", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 556, + "source_text_sha256": "8170f305dba0f51f47dbc3ef566293283f63fcb7908b27ef162022ea44d6ed51", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-bd01de209020672f6af328f029ecb58ad20fda7f9acfb876a77197ad5b445ea3", + "id": "dolly-fact-review-11172", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 11172, + "source_text_sha256": "02da06c92559fbbca8d1f4c9f11da84f6a4291c906c484d9d75ca0fa9be419ca", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-7c63396a57d5c1e7a665966912e476e35b41ed78f810ea6901b8c20433c0d994", + "id": "dolly-fact-review-1680", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 1680, + "source_text_sha256": "0f77785cece23c0fcf2edb05852946d0f03bc5c64fb71d8d478592ca839ec28f", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fed5296ea3a9859f8040095d9501b9b6e9053080e3376537f1811e0453e9792e", + "id": "dolly-fact-review-12879", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 12879, + "source_text_sha256": "ed6402022f9de852f330189871c01b10eb435c9a211116810064b93d4dbcfcf0", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b329fa832db5ad0acf308a7fc941cc188a4580fb7c0a5bea32aec842b5ca7e2b", + "id": "dolly-fact-review-8335", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 8335, + "source_text_sha256": "28d2864ff091b3273285839879808e1877b048f907e2195f5db253eb9f6629f8", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-c5a7777bdfb3d64be5e06c121ad7c637588ecabe438506b10e43419e6dab7de4", + "id": "dolly-fact-review-8166", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 8166, + "source_text_sha256": "b171d654981134477b722e19af7501b1daf01ea3b2d7b7965b814272e660bca1", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ed6e4014e6b7f4d3360b134d44ba9851b778d02afa0e3fcee1067754a9c33b61", + "id": "dolly-fact-review-3642", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 3642, + "source_text_sha256": "b2eca7e814826abe69d49530400a7aa4d5c834f8e817071cf9aad5b36c8fa2b1", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-4b09e1024723d80b09b0b8e026e00879f33e887c41f79ee35eb591118d3aca69", + "id": "dolly-fact-review-8173", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 8173, + "source_text_sha256": "3dc2cead97b58244675b2ec5271d5aa72d81bd2d616d2d0b30bf1d8a2c47acee", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8782cc14df70b0ddb2df99bbec25492e8c87a7b3833bf16fc5a56bcf2424f297", + "id": "dolly-fact-review-7459", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 7459, + "source_text_sha256": "34d652fb1a84d6d53cd2e9677de5903a6747603dd8aadec56be88a7f5ad34ec7", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-937bb041270a75ab99bfb4530c5d0f1b786a9958f5824683efafa12c82012e86", + "id": "dolly-fact-review-7000", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 7000, + "source_text_sha256": "466ed394c919a3acb397de018009b01d7d2fc1bdc639714ebbe7b38d758a2a62", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2adb3314bceca49972bb0f5f16e9f2b8518505eeb330b441607bd01ee62fe042", + "id": "dolly-fact-review-13185", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 13185, + "source_text_sha256": "a928c043ea07aa28ae76df6fd750c9c9beb95ab6e495c0813668b8a7e0d130d5", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-908fd9a89e4d4e4496cfe45a4785a58dbed1492a0358f44182471201f7f010d8", + "id": "dolly-fact-review-4819", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 4819, + "source_text_sha256": "1fc5abdc1d68c1391d81a210c970112cca89f1198712eb78ca61d1a0dac1f319", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Large/medium/small country-area thresholds not specified", + "group_id": "dolly-source-57a10198de951a7f877b155507b53ce1ffc18d992280917479b61d2f0799a7c2", + "id": "dolly-fact-review-9351", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 9351, + "source_text_sha256": "af234fa11ee8ca5c3c49af49d682fb5535ac747c50c6dcf4ab4d20d5f69b154e", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1600ee329b41172647258b952df7da4e051edc1ea548e8d9198dd3a28f9d0bf4", + "id": "dolly-fact-review-6970", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 6970, + "source_text_sha256": "6155cf57de5df594537159ef30a865287ab53278832441f59ee30efc7c5fcb90", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b8782bb38734c7d96da5f86958b4db00175c88a9375aa148bb70f452820c399c", + "id": "dolly-fact-review-820", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 820, + "source_text_sha256": "47cadeb713f40b275032df3150022de7d94f5611965ee0166a33b9edfa74bd00", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Most prolific religious leader has no objective criterion", + "group_id": "dolly-source-6e2c49dd88d275ef4b2a4f8914ad7c024dddb1338aa762e99f71aadf3544ce98", + "id": "dolly-fact-review-6223", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 6223, + "source_text_sha256": "28ad1f91a2b6418917f4b85f3b32c1982806a10e4c0b6f38e5a2e6856ac02f82", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-598862b84117be4bf10f486afb5cdd89d4a2639ad9ab77bfe193c739659e8a1f", + "id": "dolly-fact-review-1629", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 1629, + "source_text_sha256": "598862b84117be4bf10f486afb5cdd89d4a2639ad9ab77bfe193c739659e8a1f", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-4ec93b26c54a4e202641c052e2b92f74a4960daebf47d9a4c51280dca8870675", + "id": "dolly-fact-review-9471", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 9471, + "source_text_sha256": "c4c3edcc360bc57dd268af6f1480c9f1b380c44e06c246e1e4425070e34acedc", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1fadd7abc653a8e668d0611de7852cec04ce4b16f249cf3cf380499dd2a332bf", + "id": "dolly-fact-review-6358", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 6358, + "source_text_sha256": "9fe31e7f2abf32d198668c40a31ac2cb77300b1a0fb649403da63ff5fc10de6f", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-f7fcf4ac15a470ab12570bd0eeeb6babea6d51db5c5dc912eeaebf1c9d942ed3", + "id": "dolly-fact-review-13305", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 13305, + "source_text_sha256": "2d9be14b68cf8b9cd15df61837e896f1ffecc766ec469be2c29739c43ba63028", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-69b915fc0f9c0735e9a91eb5c07a5543e150ea87f58b5219a21b71d97963dd02", + "id": "dolly-fact-review-7781", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 7781, + "source_text_sha256": "898b719cd66f951dd04e5d6e4c4bfd9a725c41d90f4cb1f493400e1861f7972e", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2cbe9fbd80267489514c0ba75e5d92a96ddbd78c37843c22bae6a85141b1bb22", + "id": "dolly-fact-review-14701", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 14701, + "source_text_sha256": "5cc6546b4c2072b6e6e1e57d92552f5e7a45d69a12bca3f06eecbb6aa5d990f3", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0aa66914211054c7539a539534373613ea47d295294ba6a567ee93a8c5d00e16", + "id": "dolly-fact-review-5074", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 5074, + "source_text_sha256": "81cc6783e870c5a0fc7d5528d7671e48a1ae5894ee2309e496d55a2946ac675d", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b6f5d7a0e885f79937e697d746dbbc870c8aff13ec45e9db37d9c9bd58fee904", + "id": "dolly-fact-review-1063", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 1063, + "source_text_sha256": "c08000ac6f86ccf4ef636988a77730d2be6013135504bf513aadea5f014fd47f", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a30ba20e03c0fdde0a46045f0b3aaf4e07b3fe5e86f2062d89b2b277e3a44bd2", + "id": "dolly-fact-review-4679", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 4679, + "source_text_sha256": "a007ffa1f7ea08a795dfee239f41c4805c44e53650f37949f776ea2a6207253d", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-480ff1d0007d4b325752d5b96edc0da6a27fba06304110720d1d793f822b0158", + "id": "dolly-fact-review-12635", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Supplied passage gives Arsenal achievements but does not establish the requested comparison with all English clubs", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 12635, + "source_text_sha256": "41cc0fd28feef72084daab3c04b7330d6700978327ed8f93c5789464226a035a", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-01037db36d3cf7c6bbf6346ccaa5ecd7e1d01b67f35f29cc92ef6b991e97541d", + "id": "dolly-fact-review-6226", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 6226, + "source_text_sha256": "dee7ad1ff9d546c0a156bf9b113acee739b28eb04e9462984666970cba32ca1a", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b1808e544c75ff1a3baec1c93a036b9d1b201ba9af2624c52853ac94ce0608b5", + "id": "dolly-fact-review-13747", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 13747, + "source_text_sha256": "4d9d19f5ac4e2f03d2312a5a7c083d5b7615c9bdac821eeb8a8ae37c5acb9bb6", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a9754ca64d0cdb52a63e2c285c2e9a2a530f9b59cd6f99309e5c61d666945e78", + "id": "dolly-fact-review-4373", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 4373, + "source_text_sha256": "fc5857700d4044c5506397bced92c1c3978b8462150c6c299c7d3d4a0a530a31", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-12a1ce3d93a9b5758f9514a0b59157548734a73c624c5cbd8492fb350d62c6c0", + "id": "dolly-fact-review-9002", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 9002, + "source_text_sha256": "7eae402c746ecda83756670f17d588aeb50a30e59d5ab4835d47c459f145d62f", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ac00cce44e8d67e7de12462d825575008572bfbd176d7ec599113d2f6f1763ed", + "id": "dolly-fact-review-859", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 859, + "source_text_sha256": "b5ee409fab32edf80e1a0e4ac45219c7dea06258c8ec910fb0aa8e3f150dc7d5", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-256e073bb6badb7577102eba351e8cba82706e4a5d80575ac1e2317d245da685", + "id": "dolly-fact-review-1044", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 1044, + "source_text_sha256": "f2259b23b7203cd9052578c98c5873fc399a95fa76bf7aee6a7a7f2e18b118c7", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fb8736bb17ab62256f424960cdf3a26a2144e4b087a717d79f41c376c8541d63", + "id": "dolly-fact-review-9841", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 9841, + "source_text_sha256": "172eab39b84e603a7dea3ef41c655e508cfda54c8561a3dd724f6880238fb746", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-efd56a0733fb8e472ea85620155440f46b85d00b2647b3747c1bbaf0e5f37d94", + "id": "dolly-fact-review-4496", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 4496, + "source_text_sha256": "4eab972d4e2864f01b58deb44a9098c9e0beb490a03cd39a96e53afa7b9a6bfb", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2c842b798c0df30520ec5a90825e7e713d0d0b7a35df26a8b520b570a6006564", + "id": "dolly-fact-review-12633", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 12633, + "source_text_sha256": "b76c084b305bc45b8cefe725f6dd4c22f9eb91f9f55dc995b4e4eefc7b9c36bc", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-05cac691bac2af16036278f8a2b094c16f9f50364905a2f22bb46463f82a29fc", + "id": "dolly-fact-review-3647", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 3647, + "source_text_sha256": "17db560da570c349a10ca8dacefdf3069ecfb18125a4d57813f9dbe9a2c470a6", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2606ed64370622ec97de8a366c5226c3fc2467f0445bc332763d4ea94466d250", + "id": "dolly-fact-review-14293", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 14293, + "source_text_sha256": "1c306f0ff73c782e19bd5373bf0b9023f7c0051a21aaebe4e23d5a5f40d90ff4", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fc2ae32b07acb97eafc987bf2be8185f037ca70485f1a9aaf6bda6b95412912d", + "id": "dolly-fact-review-4277", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 4277, + "source_text_sha256": "54b7fd84778684b1cc321512a4d0bdeb9a6d2cefd2ece5c1944831241f5b7ea4", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-f09a82b830b681ddb98fa8fe10cdcd3073e85ae7b399bf53e392c60b96016b3f", + "id": "dolly-fact-review-7127", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 7127, + "source_text_sha256": "32d16d3b6b7b858ed39d283779c7508efc6b399c9f90b5180d0c686088acd2c5", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-92a29da1b753a88dec57c293e44986d578b922d42321d3f9c2150872873fe023", + "id": "dolly-fact-review-545", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 545, + "source_text_sha256": "656e30da414fd7d1c317d338036653d3a713479e4088b0ca6bd4a10498706ec6", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1ffe12b5c622a9b94389cbaed86a19991cb2271f8f19f9a483fd92f2b213861d", + "id": "dolly-fact-review-1503", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 1503, + "source_text_sha256": "004156ae2004685146579125ba0882425d3164d3a81036c4eb2e7b1069852c03", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-c89d6d8e33b45ced67a1fc82089fd191dd2733b894f0fab19e7605b7d1d2053e", + "id": "dolly-fact-review-5907", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 5907, + "source_text_sha256": "6cd76fb7830d30a4c64e968bba7f89c6e21f09f0dbb06e58248516ada09bc6cc", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Instrument-classification template family already occurs in development", + "group_id": "dolly-source-edada16b5281e14e6d4170db1b395d118b628f501c4d5d105c0141db7c32acde", + "id": "dolly-fact-review-8555", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 8555, + "source_text_sha256": "f2f47f0261afbe44d1dff56f336d53dfd40a89f7311a3ba6bb89fdc225e3188a", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-79cb91bd61fb0f923bb470b90ebe5cb775f6c8e143237f155cbe88b3cf6f7270", + "id": "dolly-fact-review-10420", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 10420, + "source_text_sha256": "a46639e2cc9cc990a37b9ca5689ff6c32b8acf2159953d3c310f2316d808e759", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-45dd156e4838ed9343edf7a9e976bb961815da31e2dbb37b5d26d16c08560762", + "id": "dolly-fact-review-10632", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 10632, + "source_text_sha256": "9a98a80021809ea06488f7c015502eed598aad5c1f4796c3acce0827fb6344d1", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-474188dfec7d2b02ab3a152367c8b9b69f875a4ca5ebabbdb95edf70e866e0f3", + "id": "dolly-fact-review-9819", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 9819, + "source_text_sha256": "016affea9c042119ba81f3ba8adec42cf37bccd7600cf96b7e881448d5eb9516", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b8c4a4b3cbdbf1778fd09597d8f114f0838b75a2ae7d0bdbbca5d31936b43a15", + "id": "dolly-fact-review-14687", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 14687, + "source_text_sha256": "2465023aca6132a55639aadfde235679ab09713e3f10183006b1fb306946818b", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-6ae9da3953d5b80c0d72c1b9208c51cbb1a235925a44a59448b862cffbeca53e", + "id": "dolly-fact-review-6819", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 6819, + "source_text_sha256": "1af21fccfd5006d6d74dd27692b666b4000bfddfd305e1f2cc666ba9b78aaaca", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a43efacb11d85044a14ac7e440494e528af6769f99e2aa74ae50e46932bf0885", + "id": "dolly-fact-review-588", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 588, + "source_text_sha256": "5518e372b048499c0f180a124cc40e84a6092193c6befa0a273f630a8853f2af", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9dc191547af9bc62e7ab900534229a56b625ed777ed393a56b14fa90ca6ccf1a", + "id": "dolly-fact-review-9639", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 9639, + "source_text_sha256": "a1e529a19e39b628fb30a637f635dd86856e465d5bfcdb3d8d1c9d7b859fb296", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-24ad48ed15c279c05f98daac9ec2eb83f82e37eab7f7a3111bbffc685365a3b6", + "id": "dolly-fact-review-11111", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 11111, + "source_text_sha256": "09763f75dea8d88fb5cf72a83a067dab0bde03d0055c3505588c6ef6dcc2af33", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-546022417905b9223524856ade2a14779e01c36039a2ece68adcdaa71a921d11", + "id": "dolly-fact-review-3711", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 3711, + "source_text_sha256": "0f9c22e6ca83b9f33999e6dbd2ce693c1d1a5ac7744d2e31ff0223e0031b1d87", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a7931f6b6ff6831bb6f998fe982218767701e18d21c87a5a4ed1cad3d9d6d6bf", + "id": "dolly-fact-review-8912", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 8912, + "source_text_sha256": "f25440ab0848e95a18d70fee3a56a716c7ca6924017c4c70292b48733c7f8870", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-4407b46c95eecff7b7475001c7c743233b5304530ca62063257904cea0ef2ad2", + "id": "dolly-fact-review-10543", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 10543, + "source_text_sha256": "cbef7476d39ae525d28e79e077546f5b3417782e117acc5818f31171f075e4fe", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-02cc6b911f1dbe341a163e60ce63c6ce865f32477a8772ced5563c6e1c623dca", + "id": "dolly-fact-review-12150", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 12150, + "source_text_sha256": "25413a3bdbe17680a13db7225cd6b2937685846e0dcee2dd51db0bced6a8bfdc", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9b5d845f5a60cd165ba761754c972d2ae7795d549421bf0bb6d99b9d5ad9cc15", + "id": "dolly-fact-review-9917", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 9917, + "source_text_sha256": "b8d4cf7e1bd5de333eb951876bdc95aad51fb6571502105ac667d52b25f13726", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-94dee453972d3492a29b332df59de07127ff3537be02ed633c17c8dbf1a7ba42", + "id": "dolly-fact-review-4494", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 4494, + "source_text_sha256": "e2322d1cc4e255e5ccb35dd0af44c3dcfe9ab0ea6fd6a9587cfff6f91f0f85dc", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-83c1af479e6d5dd4b3e1e27df2406e5cfb1a5dd2581da97729423ec04425dbe3", + "id": "dolly-fact-review-7665", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 7665, + "source_text_sha256": "55a9d0035fa7c79c9e0ad216a48282222301032cbdd808dead0f3007e1cab173", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-69c84a7d1eb4b1c9c34ee40158cffa0dcfc4fdfb4af0c3ed36c9eb92ceec7db1", + "id": "dolly-fact-review-6329", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 6329, + "source_text_sha256": "a4f7e2ee9f52439fe225322d74153c7f0eb8c3e33208b65081dfed367fbfa7e6", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fba3b0ef51a86e7793d8c948b00d1414178814e9756516e0d3e5cd99da120a0b", + "id": "dolly-fact-review-3947", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 3947, + "source_text_sha256": "59aa3344af4ede7c4ff98cd0f909030da78697a4e36e12162bb4a72c783a74cf", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-5f7cbfce32fcfaccdad881128fb1c632e33826334c6ff77acea23bee1380dc6b", + "id": "dolly-fact-review-4429", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 4429, + "source_text_sha256": "4aa12f86dc1b1a5e35c6c39714c35ded41f5685a145d7459ac278df516913bd0", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-df53122a16ddd28f5dafb8d0a20afeb1fc7fc19c88e02776eefd086c938d5ea1", + "id": "dolly-fact-review-7664", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 7664, + "source_text_sha256": "af106267d1f04b66fd5fdf6baa9e6a58eae256dc68a5fc2dbcc5c429cfa9c1f8", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Fish-choice template family already occurs in training", + "group_id": "dolly-source-a21cc7ed7326fff6fb6beb706dc3fdd469f76b2a4a0d9805ae75eae74a3adc6b", + "id": "dolly-fact-review-1945", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 1945, + "source_text_sha256": "d1f6ca40efe91b4407f681ec61ea6d89d4a93540acb340b6c5eed2087d7e1cbf", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a3dae66b1ff0c5e4efabc4f85f61c760a271f90ca47f40b7fedde7fd713c186c", + "id": "dolly-fact-review-7644", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 7644, + "source_text_sha256": "ef149c472bc35a9b80fb1388608d5988c140329fc7d3dbe08cb5a12cc171f999", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-52ddb06133337a926b074f605af528e6c75a4c40ff48b72c213e0325b2a4cfc6", + "id": "dolly-fact-review-4006", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 4006, + "source_text_sha256": "388b46a29c084abc87746fc3ff67368dc2e64fcc2aa33fefbd053b1f6721dfca", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0e9a46dfc8cd2a9c20c667d1282e76a91307c1a6065459bf356d7371e45b9899", + "id": "dolly-fact-review-13481", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 13481, + "source_text_sha256": "490c6dcc8e8e747b33bfbe34ca40c7ccb4ae99e7b0d45edbe02c975b2929299b", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9aa5b731686ed58e3d84821f0505c9ec6ec24801120b8a9172f599f367ee32c6", + "id": "dolly-fact-review-7909", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 7909, + "source_text_sha256": "f76dc3a5e941cd7565c853ad04a879a965db461d996363027b5d17225ebc0901", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-84dc293172c92a2acad847927eee0bad9c4316ee0723310e0c09a11d178eecc4", + "id": "dolly-fact-review-3113", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 3113, + "source_text_sha256": "98e94b89558bcfe96b1ee62000fe61b22d497c9db3cd3296f9eb40d2cf33c68d", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-c55a5fdbb7138976ed240d1033251b0893acacc24c138afac96aff87f00d3400", + "id": "dolly-fact-review-13753", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 13753, + "source_text_sha256": "06b1ae4653d3a1a9be2cf058df3906e95c9cedf789bcb48193aadc1462ead747", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-3e71918669d88e419ba2c83080d894707e5c2753b4a56591ae7aa90cf78596d2", + "id": "dolly-fact-review-13932", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 13932, + "source_text_sha256": "f85a99c6af4e638c43972f2d3252289be6f6ac84dfbd5f98a963de313491c781", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-d6db9e39418ee0137faa34e8703853ebb1691c4f7d4b6365d04ddf2ee29cf40c", + "id": "dolly-fact-review-168", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 168, + "source_text_sha256": "304a38dfcf368a0da17414ed77653557eccfaecc488265d23d67fa1774615f94", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-42128404b0c6fcbc8965d10831d1a2ecf31e1431f8cb5e7e251ef9f43fbdecab", + "id": "dolly-fact-review-14066", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 14066, + "source_text_sha256": "0759ac1563f09e5963d188280fae1d5d00781aec92a0db0141169b4fc7bacbea", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-284207333007922f135ef01616949854f5bf6802e5358bf04e01383e0ff6794e", + "id": "dolly-fact-review-11265", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 11265, + "source_text_sha256": "c5831908b618200cdf324bd68b2e9e4bcb52c7a4ae688bab20b3e25e3d0ce9f8", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Former chairman unnamed; ambiguous target person", + "group_id": "dolly-source-ef60b884b3b77f49d174511f80535a84a32d07018f54f32609f2c8fb4e3bd354", + "id": "dolly-fact-review-192", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 192, + "source_text_sha256": "0c7f597c6402e336efd8067d191184a90cda0e147591b3416f1027eb60caaedd", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-f0098906834c6c3fbd976b1319ef3ce9bc39edf54a1a6c07af352189ca1e05b8", + "id": "dolly-fact-review-2344", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 2344, + "source_text_sha256": "1a14dba2d952a4eeff759131acb940bdf05c0946a9341d09a30a102fa41d74dd", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-811e84bf96c177a11a3db68c3a5721fca2d115ac7c46ec91de30d91f6bd8b810", + "id": "dolly-fact-review-7658", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 7658, + "source_text_sha256": "53e3c8f3a3a7e25901c0f9ef79798b7dcc1d6f6cad7a25fba62fa3f254cdc510", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0a8a13819bb25c8d6c02ba8712195d099212f3b0ff0191eaccc37dc97cacd58b", + "id": "dolly-fact-review-865", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 865, + "source_text_sha256": "6914ef8fc8e8ade7b34861bd2faaced7c8737265bd42f787f60b03af2cb1f254", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a2e653c51c041560408ffd9b7bb1d0401521e2aae03e5442ecd95e92e1b5cfcb", + "id": "dolly-fact-review-11542", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 11542, + "source_text_sha256": "263e5795f3b8953e0935979976a14547fd790b572910b1050456190a23c2208a", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-198ad8d0504ffdd4c05b164d5e05aeae794076da6c145a66bf89dd317f2c7581", + "id": "dolly-fact-review-10391", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 10391, + "source_text_sha256": "a364f0fd32c07de980468d034f8ec71eabe3f988c939b2359b4ae0778f1779af", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-00311b12d559ff7acbb027d4abeb47972d22abca5f820a99774c68fa25e5cc67", + "id": "dolly-fact-review-3628", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 3628, + "source_text_sha256": "ba74f778693eb4e16aac546bc606323b717e9a74dc06acd77b6fd7ff5ad603bf", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-908daa4665415eb458e6bce1fc521b39e6aaa8161a3c60e736bc3e91fa4d7108", + "id": "dolly-fact-review-9949", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 9949, + "source_text_sha256": "69fd50a0f9c12233e129dafa638160de891bcf5ec97f1a0d606a0307de7c1ca9", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-3f07a1e64bccf8e5856585eb66af0e9a73720af82389e47d1d8504b43400b16f", + "id": "dolly-fact-review-7635", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 7635, + "source_text_sha256": "41dc6f5d94f1f62e9a2014c5eb39fe38e8e5517b5422410fc36f0d31edb6e9d0", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-80f3ceb5a97c45fd7fd04b81de0f9b3742d95b391f999c1d3b68741bfaf4c3ce", + "id": "dolly-fact-review-761", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 761, + "source_text_sha256": "3a6b2b4e3229728c2a4656c7ceee1fd3c3e5dde702af59368c7658fc32a3d15e", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fbb2a47955f120b55cef7f4d1b2b48d0658a57ad8814ad19524c58f30c3ce925", + "id": "dolly-fact-review-11821", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 11821, + "source_text_sha256": "f2edbfb30c54dc73afb44c5a4f245ac32078596aad6ae76ff656ea4bd93d611d", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-cf375f003cd1fa1837edd85ddee3ad2ac68f66a491ecf17828cb00b2b562ac67", + "id": "dolly-fact-review-4659", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 4659, + "source_text_sha256": "679fb5a8040956664ca2d8b630301a1df0b0921c8c5ccae5502941106bbb7aa3", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e5b2c224071b561c7ae4bb007effc0fee96e335443450ef1d985a8a2d74ea293", + "id": "dolly-fact-review-4578", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 4578, + "source_text_sha256": "22955cf1ed8449713f270d5623b187459b501a3de4d0f21537e01164ce318430", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ca23532742d26301da97fc8c4b0c21f0ec61a82191ab93d2d37d5708798db723", + "id": "dolly-fact-review-4405", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 4405, + "source_text_sha256": "8555cf03cfdae4d8be3ca7d95996657005db4c7b287f1cab3f95afda8280a146", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-47125b66b46b7406667ecfd6ed0c751924f33a1e3bd0b1a627716ab8c6e8f051", + "id": "dolly-fact-review-9063", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 9063, + "source_text_sha256": "73c0d09fea00c5aac1e73ca7921df57ea2087707a2bb0174f04da3b75021e8c7", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1b801b62bba5328ac8b31f96125557db1f9dd0d8ed9e2edc3ab397f1450af304", + "id": "dolly-fact-review-11323", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 11323, + "source_text_sha256": "ead56c7c9f97439facebcf8975a97920a41be7714718bed36d77f75f470f2ccf", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fec92c04ed4585c8e42023b18a761477d4b2305fa549a06d0e03e59def3f149a", + "id": "dolly-fact-review-7675", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 7675, + "source_text_sha256": "a05572a5afe7234d473c40368f44a0c16848f1b482d50a82febd9919c7ab26dc", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-40ccc41ab01cb5dbfecdac7d18250a2dbb4f3163ad40ecad608e2990a944f79c", + "id": "dolly-fact-review-2100", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 2100, + "source_text_sha256": "20d782ac032723f4d43625a84273cb4c032e4f8d9b478ab45f7e1990a4d71b26", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-b88e5c01657ad2651fac1ad24da40eb318adc3b6165160da6768f9a641b164e9", + "id": "dolly-fact-review-14752", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 14752, + "source_text_sha256": "d229c33681260c29b820b15c6829b5296f383cacf9baecf579c0a200dd507f76", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e009d445c45d8d4917883c8fe4f7fc5f29d244854718353b9adc23ff87b51c17", + "id": "dolly-fact-review-8485", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 8485, + "source_text_sha256": "b239459840f73ebe7fd67decb7d3645db1f0f24c828ee534c2f588179c508777", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-32a34f99dff14fbd03c966a8c289e0e2a614f400dde2c713cd0bc015c4e4913f", + "id": "dolly-fact-review-8966", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 8966, + "source_text_sha256": "32a34f99dff14fbd03c966a8c289e0e2a614f400dde2c713cd0bc015c4e4913f", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8dd5e7ee1c8d668c83b977f8921b5ed2ff09126eee2d60f8b14afe5abfdd8866", + "id": "dolly-fact-review-507", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 507, + "source_text_sha256": "9b2395ce6ab2de3def3b59b438a0dde35d3c7750cf759600bbc1e6c81ebfa970", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e55e1a6e8e0a71818f4e597478649e182aae80105cf24dc07a618d07b1763bb1", + "id": "dolly-fact-review-8667", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 8667, + "source_text_sha256": "2f8f7771afcc3b7d62b7c1055cb7ca8c5e54b49296bc8493ec15e8029501c05a", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0fa9b178605131a2667bee85586a77d4236cddf3ef031aec769cec108496141f", + "id": "dolly-fact-review-11911", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 11911, + "source_text_sha256": "e5f4c2cb26f37591c030d37f33e3332c32c47fe2a24385afe27d8fe3cd6c3560", + "split": "train" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Missing city options and undefined monument comparison", + "group_id": "dolly-source-b44dfaa9bccb69c95dc69384f3ab5e056abafc33cfe6464078a9655bff2489b1", + "id": "dolly-fact-review-8888", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "open_qa", + "source_row": 8888, + "source_text_sha256": "a5049b31bd841dd973544bbf1ca7e72bbfb471e9629a541ddedb6f9018be648c", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Spicy/sometimes-spicy thresholds depend on preparation", + "group_id": "dolly-source-08add5ef1d47fcdb5f0d9d5b1b0dc298f90c0ec573b9a8f07fddfed0a3cb9ee6", + "id": "dolly-fact-review-453", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 453, + "source_text_sha256": "0cf42cfc852cb8b1e8e07f434e819236931e2550d2df5123ebf90937ac6d587d", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-04593b3ef446d1e7f7132c0cd8f853ba836cb87084abd054de077f927fcdc164", + "id": "dolly-fact-review-7826", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 7826, + "source_text_sha256": "ac61acc66365ecbd7c244573e2185041a6217d6c0cf703f7ebf393c2e541a342", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Instrument-classification template family already occurs in development", + "group_id": "dolly-source-68edae7747d07f6747f5f866c5803297c98a305089bf6b81229b4ba9fa4de94a", + "id": "dolly-fact-review-12453", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 12453, + "source_text_sha256": "d4977e07abbe6cfbec4fa66587d4d22a5810086a1f030d7dc8d6c4f3225c5807", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-c2a9dfd008fb28fc79039c686666cd6a5b5551959dd1a22ba13b749fcdb97bcd", + "id": "dolly-fact-review-13197", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 13197, + "source_text_sha256": "7bac53e815a9e028053c62e21e18d84e31449cde4a48f35eaedeb19ac46489a7", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-343a73777504c74ed8aea5a659833378b4a5d2f7b75085725a3fee8c9780b768", + "id": "dolly-fact-review-4732", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 4732, + "source_text_sha256": "5f6749751b79dc5fc0aaf68f8e539e12de9385e45dee7089edafc8a24bcb1cd5", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-993649164abf94e69270399e512c03a09c3424dc6fc2dcb8b806a0300a84a4c6", + "id": "dolly-fact-review-8967", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 8967, + "source_text_sha256": "78c3fcbd69310493628a762aec6d94f4a96cac42ed6bd533857b45ff98f02096", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-436c0deeae876239b0e06889af02e48ecd41dbec70f733890a4507a9bf64a415", + "id": "dolly-fact-review-4367", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 4367, + "source_text_sha256": "53c6ea424f84ab09a79c3367bb9a3d9d4fe86ddbaf2a306d3417d8629db51644", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ea991d2f1116afc6c3a349603737c39a61c4ec8e89c8cf3311b49775586866f0", + "id": "dolly-fact-review-11500", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 11500, + "source_text_sha256": "bc3fc6b5221a273f73f49fc146b6292fb9ccdcbc6db2493bda881084fb4bb79a", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-0093becc973229e44abb62c67c1753ef2968dcfe6865b4d17d66094c7aec87e5", + "id": "dolly-fact-review-100", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 100, + "source_text_sha256": "72c30b1b3123f6adb3b60dd84025b2b93e4fd541ace610b7e6913896275523df", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ebcdc5a85b678049a971c253d036dbed601987992ec1ac7311fcb309f6c13c52", + "id": "dolly-fact-review-4853", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 4853, + "source_text_sha256": "6693f6c4615a92a484b2a87f6f18e6de8a0ae1afa1c9c3404bb10d040776ddbb", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-587e788ab0b55f638834895fe1c44256aab8150abd33495e779bc49eabe20a6f", + "id": "dolly-fact-review-4753", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 4753, + "source_text_sha256": "72a2e088faeb2401efa938eec037080bd6248d1b1e326153d6b494c41b9c7140", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-14405486bc1e25ac3e5ab68d60d48c6901c42ad7464ddc153ea422e1d09fff85", + "id": "dolly-fact-review-9129", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 9129, + "source_text_sha256": "d77bb5b06ecaf02dbf43445d319d82820ccdd696a775a41d5722b9bc67721127", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-aee937563eba9c1d0e92f0c65a21c6348e3dc44987500571ceb3729378b4fb84", + "id": "dolly-fact-review-3331", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 3331, + "source_text_sha256": "430df747fe47a1ca5f7938b4e45c2b544b6a288be78baf3c3e907757c116187f", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8e2fa9888537744976bbfbbd364a186f8613b72074f60806a6908bf5cec1cadb", + "id": "dolly-fact-review-7552", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 7552, + "source_text_sha256": "e89658a7b9dc424d31b783a299c4d290136151f2965ff6d37322955de0f3781e", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e496ff7bb59e59506ab577558a361de83a22404d881ec0eb5e7c317782d4c164", + "id": "dolly-fact-review-5621", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 5621, + "source_text_sha256": "c43e3a6dbe07d03e6d9332a24fdf8252ce9e1828087a4835341a45a5af651519", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-cc26137a37e3855d188ac889280354b1fce771ad676be83b52f759289647eb47", + "id": "dolly-fact-review-12912", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 12912, + "source_text_sha256": "e96449c0d0bf094ffbc1de18b91585c57a99e90d9d72f4182ee860606fcbf51a", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-d44068e4d331cbf3be84ffd1ffed7d0a3aa6ca4213d5b32b8cbd078ff26fb31e", + "id": "dolly-fact-review-8262", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 8262, + "source_text_sha256": "bacc7e299f6772116f939b5307176e5fa837ae4af84c88d65d37c2ba2440c871", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-6ee0af20d4225fc3f24ee20089467f2bdd64c8080ab04c2adf7d58f402005285", + "id": "dolly-fact-review-7797", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 7797, + "source_text_sha256": "8668ae91c5e3c6a5eb54e3bf0dc6144e9af0f523696d2d459da092bc1186ba21", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-7b8f68058f89fe2651f3bd7cc69aa57447dac266450553995697149550555fed", + "id": "dolly-fact-review-9890", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 9890, + "source_text_sha256": "b3fde26492ba6f9a7bfbfadb06891a96c80bceef7b830046e289467cc1340cde", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a6f07730a229318139ee024a268a96a5aa9ba5234a2b9c05fb5d0d383bbce7dd", + "id": "dolly-fact-review-12882", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 12882, + "source_text_sha256": "80fde6232919f2be1fe3d32717f0ad82a0174e37a66f77c63203302eab7346da", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-73a4cc859f857272b37a49280a1aef08a812a4f952ad1f9a42740c4cf8778b6a", + "id": "dolly-fact-review-10133", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 10133, + "source_text_sha256": "47ec1ac505c17703b5b9ae6d6cfafdae0423c901bc9f5363c326929b84942174", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-7e665bcff46490cff8ef7e0270bdc617bd3b5e8408bbce593c2b68a896d8d664", + "id": "dolly-fact-review-1126", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 1126, + "source_text_sha256": "09f0c227a27c5823012e359bc38165ce0b923a84d5191e815128ea00aa718295", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-64516965be4d2639f025d7eabf2bf5904c1c4b643ff353b968af0a61135f51eb", + "id": "dolly-fact-review-10203", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 10203, + "source_text_sha256": "8220f1a09a436636d015b2781765dbd48ab8cac160e120214f634919ac2a44df", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-058ce567bf9ce384d447f75208b8c44429283d300c94b9d5a53a02564442348c", + "id": "dolly-fact-review-4094", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 4094, + "source_text_sha256": "8dc3d72e095a4583c2109ca723aae702aff1ac3ba145a0763d05fb4cf6f727ee", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-61e2ddfa844b0f4b9893b16ba09a0c5b444d6d77518641c221cb08caefce0479", + "id": "dolly-fact-review-10267", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "classification", + "source_row": 10267, + "source_text_sha256": "7ac9391c7933d7eff7e1d2de5acf79253d30477c264d7498a115b8d709a7effc", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1af0f3bcd947055aa7355ae5fd5dcede0f4d968ebf3abb053c5d268749577a1a", + "id": "dolly-fact-review-364", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 364, + "source_text_sha256": "dc727c502414d3ae700f679bb893fed3ac8e96d647c74545f3da16538cd8cc5d", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e1fb734c8693e0e5d759b94b9f3511642f07063c8365734750451c6f2837db1c", + "id": "dolly-fact-review-2302", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 2302, + "source_text_sha256": "2eaef0b051fa5ef9047ad1005b505a1a550607e54078f78100189783306efab9", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1fb1423afba81cb76919993bfb2659fba41c47442a851cf739e48d8c4f68a837", + "id": "dolly-fact-review-1382", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 1382, + "source_text_sha256": "a37a19f7863b5967b54e7bd9eb165123f8bd63a5d5bd0ac52da716b3e25a3c41", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-249b390275b8f0d1f224c5bae3f1e4ba550fb5f8d54445b61f7a11c89a43a29e", + "id": "dolly-fact-review-7573", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 7573, + "source_text_sha256": "6d71d7bdbc07fbe4a277a8f8c745a4e60e8331c35360b8099ab23e569c275913", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-9e5269e59542d91c7b2436e998a81d63e6c17e0bd728d8ea852f18953222ef40", + "id": "dolly-fact-review-1835", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 1835, + "source_text_sha256": "a0532317dba662ce8eb8960f3a6a5830ebe0feda1f8b9922caae540334207f91", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-26a0d8bd501f50bc12d7e0ed742350d2b173335cb532b9c955fbe132191ae9bf", + "id": "dolly-fact-review-2534", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 2534, + "source_text_sha256": "e7da1915aab9379e62ee910dea94ef29a0574dbf6f04eac766671457fde4bdc2", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-ffce8fe8b18f9888145097a3b86098add183511338a4bd2b61f8cd382031e346", + "id": "dolly-fact-review-3799", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 3799, + "source_text_sha256": "98b3c854fe9937e8c2585f7b68fb11b552c25eaf15dda22bc0b19f3bc7d51ddb", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-5c4db3900a655f22e54886baaabaf03d95ce990d07c7b769464ebc24ca93b303", + "id": "dolly-fact-review-656", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 656, + "source_text_sha256": "54ffa402e847c62941f6596392f5c6bca1f1a86679a7d4bbe8980ed47620a318", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "No particular polygon specified", + "group_id": "dolly-source-c9a1fe4f7a0c9a3f2ca4a3432327132f00c006603cd72aebb5d5c6374da2f910", + "id": "dolly-fact-review-12063", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 12063, + "source_text_sha256": "62e9c7839368ea7fd6cc8d801d2f557a8d7abdfcd029c93af36649ae6f48eb71", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-08c11940c54df50adc0742599f85999b08dcdd462fe9a5e737fd6f12915c408d", + "id": "dolly-fact-review-1218", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 1218, + "source_text_sha256": "1f8f40887d912973e9e038d8fcf643b43f39e2f5aaaca05e50cecd54ac41ef4e", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-03d20f1f01caf14d35f256dd3d731cb9d5f122ab3c275712e35dcf7b418dbd88", + "id": "dolly-fact-review-4951", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 4951, + "source_text_sha256": "f7c3676223749a426eb3db4b917c63bd1a57013388a83f39f4e87bf24e926d42", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-010b051ec9e7d91dc4805bf8891189ec469f2fc89be9a44f614e6d7e8de6a9d4", + "id": "dolly-fact-review-4810", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 4810, + "source_text_sha256": "5495eef35ce769346a8c1cd28ae155f0b6a0343df03daa4b15eae9fa6d119e54", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a18f9fe019bdf4b4e3970798080037a28abe3a99b07b4bfccc302ec464569e9b", + "id": "dolly-fact-review-4516", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 4516, + "source_text_sha256": "4bc687514bc42d66fd7a9139fa5a63d2fb6f7d39d08357a514e5a0ad6c8e6975", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-26fb1b6f97d6c4499f8e3642b183ed42c25820c244a29876ee825563285c3f18", + "id": "dolly-fact-review-1572", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "brainstorming", + "source_row": 1572, + "source_text_sha256": "e242b73322075d1fee84a780c08801c196952651e28c19b77b95e56b36497359", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-48c052059d40942d91e5fd2b2ced34d5d4a760f141d02d8700fe64533b716f79", + "id": "dolly-fact-review-12799", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 12799, + "source_text_sha256": "e2869069ef2c62ddfa95aef77e14d2183f696929b59e3bf4b1da6436c786c76e", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-95d5d67d9b34e55128a909444c56479af5ec81183355fde1a3eaa694b1841bef", + "id": "dolly-fact-review-8399", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 8399, + "source_text_sha256": "325bb64cbed9d41c524c505432714c2e0483d327c7170355161085116881a30f", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-42513855505dc7887a8ab8b393a3459704062940a3e397ce06434d9d9650b7ba", + "id": "dolly-fact-review-2773", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 2773, + "source_text_sha256": "20f7c75666b27c0edee852825fa8f67afc04f7c2a78ccc7357c831377d843e90", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-1d4cdf56c58fafa8e2b6a7be9d6805d5d354415d1dc8005405c89974a728abf6", + "id": "dolly-fact-review-10488", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 10488, + "source_text_sha256": "0cc60c8ce2d2e7eafc6948852bb0c6ca1b1939fb0e57cbd150d152963c201ff7", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Sport unspecified; referee duties ambiguous", + "group_id": "dolly-source-ad71ec69dc8832c0a17e629d22cd2e3f47fbcd96f13d2e513e0d346221479f66", + "id": "dolly-fact-review-13898", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "general_qa", + "source_row": 13898, + "source_text_sha256": "f5cc4c110fee3ae23a55c211eb6453d0d948b570fc437316cec51f6b32cb2102", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-217177324a315449055f01d097e691c5f25cdcb6ef169d96982c21e9d64db193", + "id": "dolly-fact-review-1821", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 1821, + "source_text_sha256": "2f9840b8ecf959cc118cac099bcdf5cc1c1027c0d3338eba02dbcbbfc1eb04d7", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8e491aa91d73e4ac287308ef4fba2e951cc7d87d5041a559f93929edb3a363be", + "id": "dolly-fact-review-4381", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 4381, + "source_text_sha256": "ed54d3996f3904cd9792de0a3e39d8c70e1389f68d719688c40f693b25bfd3dd", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-38ee6aa578c853ac9959303a2ff1673884c4fbf530bdbbcc0bd7a03f5f89558f", + "id": "dolly-fact-review-5511", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 5511, + "source_text_sha256": "363ce77aa061ef42c850fc3951d2482b72f3c5ceb609d03e7d3bbfd377701b95", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-daf7b59160dacb24ab875a3b35bafe5b60da97fa0329cdae0fe6bae8484dbf36", + "id": "dolly-fact-review-6366", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 6366, + "source_text_sha256": "322831a4bae28682cedcf5684420c3e6be42a799dfe61c8ecd04f6e924aced69", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-618f82ad96b47daf5e643c0d9fbfe132c2d8ac8681d0a14f07615a63fbe0dae1", + "id": "dolly-fact-review-8401", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 8401, + "source_text_sha256": "4e0180bb944ddfd0d7d8d85697744f4988023ea50218c3732153966b708c8c1a", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-32df365934031976619b2976f9f48d213d4d8a64c9eefb3a25d3fb54247a565e", + "id": "dolly-fact-review-12786", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 12786, + "source_text_sha256": "b76ab831d949f5a76f95898546abd9597dc0899b06e707050d74dc9fa35ae75d", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-08187268189e21cb18aa173ddb3b0d364971cad2f338c2458949ced2e3572177", + "id": "dolly-fact-review-10483", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 10483, + "source_text_sha256": "1174195647b92bc67dea7b99ea32ea392d1433b90d678d5531b54d8f66b5ad61", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e01def8920b5c10ca7d979c8885119d4eb28b48b57b82ca2068760905d1d4fc5", + "id": "dolly-fact-review-12688", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 12688, + "source_text_sha256": "667d8b218452e0d46468d56ec1c14ac30cce66261bba098e4eb8c422dd839295", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Normative policy choice depends on unspecified factual and value criteria", + "group_id": "dolly-source-5ded3192104e84b778ff069795ffc33e7ea40b31c93a1f90b433f47f59b98d50", + "id": "dolly-fact-review-8154", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 8154, + "source_text_sha256": "55a8c254a51d26b2b0df7d8ba19e985b1414e1056472330782c61b285f97de70", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-2ac10d8fb190e7ebec95021bdc22fd04666ef3c4fdcc8c564fd1a99db4103ce0", + "id": "dolly-fact-review-2584", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 2584, + "source_text_sha256": "2dca9089520cc871148ff92ae29207bbbb7796400df4c6d4de507eee33a76aef", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-02efcb5731a2ba9fecd8f1f360871c995f8bc37a894dac264dac641f22c84ce7", + "id": "dolly-fact-review-6786", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 6786, + "source_text_sha256": "43875f182c3d6704e018bfa8e83d19fda9fc3a6f256326237caaacfc13646b0c", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-efba33e3f188ac74f4f8800fe85238eddad21e35ab6458efdf9e37e691027386", + "id": "dolly-fact-review-202", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Passage describes each empire separately but does not state the requested historical relationship", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 202, + "source_text_sha256": "c10db9983dfbbda447bdb01d967d7758f97de70eae66445dd0c684f4612e13de", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-a36ccc30283a72d213686f0cf61670b1d96d7c9c2937febe9d7a83fd17c5a9be", + "id": "dolly-fact-review-7585", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 7585, + "source_text_sha256": "3d277954d153f96e61b8683c387eaa2f8769599460957aa38cdb2a36e1bbb6ce", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-d20185e9761ef330427083be97cd2968e36189bc854e5da11a7394b296eb2c20", + "id": "dolly-fact-review-1274", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "information_extraction", + "source_row": 1274, + "source_text_sha256": "02188c81781855ae47b196a658647574c1a659884a59223d27a6b79172054b06", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-49f22ca13387a946a8de52d02c172bb72bae838d9b39fc7a0d4cb7afa6750823", + "id": "dolly-fact-review-14168", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 14168, + "source_text_sha256": "90ab8d2503428e203c11cbb2c1ac8270bf1870f50c104af9fdcd273356ed23b8", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-805c713e76b8cca95e79bcd47fef50e7975ed053be47e997ee397be7805a18c7", + "id": "dolly-fact-review-7556", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Creative, subjective, procedural, or self-contained computation task", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 7556, + "source_text_sha256": "84e795a604b21ca245b7fa6745be466f378c49fbf2b9a91741c4a53483eefbda", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-27b8658e2f2d2499d2e892ea62624870aba38e575e0d6d31b98ed5b3dccfc199", + "id": "dolly-fact-review-11822", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 11822, + "source_text_sha256": "2139804736cdc05ba8e32b9e104030f95e2a265bc5cbd8887d89ba2773e22934", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-4e83365e47430f42682fbe9632da2fead39b2345e099b9b62a7281802c53931b", + "id": "dolly-fact-review-11846", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 11846, + "source_text_sha256": "8e42de8b358287625b34815a158f0345e6c2b2a1920703a8b13ce82b625e055c", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Persuasive essay may or may not require historical evidence; annotation policy ambiguous", + "group_id": "dolly-source-6dc583d226c5d9bff1d1c217790a7729efdc48b3a3ee2882ee49f15b894ce91b", + "id": "dolly-fact-review-888", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 888, + "source_text_sha256": "3502b239e6c0cdb41e2ad65afa96c014214d82db8f6cd8252210951f3a81c047", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-e6a7dc32443b612844ca4a637049bac386eb19c411a546bdd575f4ff605a234e", + "id": "dolly-fact-review-9397", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 9397, + "source_text_sha256": "0cc9087395f8fa92c7d6facc616121ff8eb57725184b39363b54dbbca2749900", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8e96e3afbc27b26987d9afb91c501b8ed2cf4564dd1f73f19f1cbb825f2c660b", + "id": "dolly-fact-review-11199", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "closed_qa", + "source_row": 11199, + "source_text_sha256": "972c29e061edd73f45a58fecb73fac95712f7766db650a70571c1bb178938137", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-77c7fb4702b1dd8d6e0bec02739f1bd6f4acb8cc89cf7426e955492abeeda194", + "id": "dolly-fact-review-6489", + "label": "FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 6489, + "source_text_sha256": "130aa1f92c59393641955bf77a416e6f13358dfd170c8caac1e2f5d892e5a10b", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Same Landscape Arch source topic and overlapping passage as development row176; excluded to prevent source leakage", + "group_id": "dolly-source-c3ae2d9e7b87f9d1ae82c85b1761f90ace5aaad5e08a9b065d04c8169e054df2", + "id": "dolly-fact-review-8592", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 8592, + "source_text_sha256": "b936ef1a6788a907e7c13d3464175180425515a544167e84cb19d118337964be", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "exclusion_reason": "Underspecified essay can be factual or fictional", + "group_id": "dolly-source-21a362c608c57192819b5c4735d1b34370a85eec9befc873aa3e2f1bc5a435f1", + "id": "dolly-fact-review-5814", + "label": null, + "language": "en", + "length_bucket": "short", + "review_rationale": "Specific externally grounded entities, definitions, comparisons, current recommendations, or factual explanations are required", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "creative_writing", + "source_row": 5814, + "source_text_sha256": "21a362c608c57192819b5c4735d1b34370a85eec9befc873aa3e2f1bc5a435f1", + "split": "dev" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-8089efb05ff8a1878ccfa8ef646591e3438c96433760c26d389a6d42219eacc0", + "id": "dolly-fact-review-7394", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 7394, + "source_text_sha256": "cc359f0c0b88b37a9edf2705eb4de1e2f41194834f0b5cb287f24a94ab607203", + "split": "test" + }, + { + "annotation_method": "Individually read complete instruction plus context before predictions; source category is not the task label", + "group_id": "dolly-source-fb85ff1e24ada519187c9829dedc589851ffc2ef5ad255bad326ecb459a5c15a", + "id": "dolly-fact-review-8530", + "label": "NO_FACT_CHECK_NEEDED", + "language": "en", + "length_bucket": "short", + "review_rationale": "The supplied passage contains the answer or requested material; no external check requested", + "reviewer": "assistant", + "source": "dolly-human-reviewed-v2", + "source_category": "summarization", + "source_row": 8530, + "source_text_sha256": "22df699a56be2792dc064fa56ceba04c741e1bb7988bc6713413269bf18fe400", + "split": "test" + } + ] +} diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/source-files.json b/reproducibility/src/training/model_classifier/sequence_repair/data/source-files.json new file mode 100644 index 0000000000000000000000000000000000000000..d8fca02cee733b94814f01e28ca3c3621afdb760 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/source-files.json @@ -0,0 +1,175 @@ +[ + { + "repo": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "license": "apache-2.0", + "file": "README.md", + "bytes": 13659, + "sha256": "3ac54f63eec84100c155571945da01c7c18cd8ec97ea94639e2b94a6860e8573" + }, + { + "repo": "CohereLabs/aya_dataset", + "revision": "f9ea04583f02a8f86404ff6c58bf75fe637df8a2", + "license": "apache-2.0", + "file": "data/train-00000-of-00001.parquet", + "bytes": 137195800, + "sha256": "51baa85043b569ba117f41516b32d1b0e4e2647fb203a08933f98a556fb1fb06" + }, + { + "repo": "poloclub/diffusiondb", + "revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "license": "cc0-1.0", + "file": "README.md", + "bytes": 24993, + "sha256": "d4e51a582f661d61647fbd74c7974b8dc1b59a47fed20aa5a9597100875dd38e" + }, + { + "repo": "poloclub/diffusiondb", + "revision": "fb620fbe49fa4420e0734bd9c0df11f51176b61f", + "license": "cc0-1.0", + "file": "metadata.parquet", + "bytes": 194548652, + "sha256": "eecd341187bc91c07f5994ad0660d40228ea025616fd57a509bef8323677c68f" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "README.md", + "bytes": 53310, + "sha256": "80e44287371de876f7b1dbe613494649ace41082e9a7caff8deb96b9b26421d3" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "en/dev-00000-of-00001.parquet", + "bytes": 90574, + "sha256": "86e985fb70b49130d751c7c5b255767348a8590793dc794ff0f34f1b8bd2d16f" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "en/test-00000-of-00001.parquet", + "bytes": 3910925, + "sha256": "5baec81ba97243728cc9557b2cba2f03aaeedf8232473db9f20cf995c6b0a1a9" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "zh/dev-00000-of-00001.parquet", + "bytes": 84714, + "sha256": "6c1dc5daf941264c1de302acf9088c7ed436d25c22e311d583271c81e7c4056e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "zh/test-00000-of-00001.parquet", + "bytes": 3618771, + "sha256": "e1ba0c287b650043b8686fa5c506d1f2718b098a7f88f724ade55a76f0062ae7" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "es/dev-00000-of-00001.parquet", + "bytes": 97926, + "sha256": "f4c4420525ef97f7a88da1bf6d8169c0faa70f49de6c779c04db94e8f278fac1" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "es/test-00000-of-00001.parquet", + "bytes": 4327396, + "sha256": "efbcc4ce1cf14f2f43f7b05be49a9a1153b2893c59d2f660c3bb58679801dccb" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "fr/dev-00000-of-00001.parquet", + "bytes": 100882, + "sha256": "bb3dd1fdd13509b0c8342af49470bbceca9cae2bb32f6e3f589de93aa4c2a82e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "fr/test-00000-of-00001.parquet", + "bytes": 4550854, + "sha256": "e602201459c29c5dbacf9a3d80cfe2524496f8e0c0b51b6326ef4205bad3405e" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "de/dev-00000-of-00001.parquet", + "bytes": 100678, + "sha256": "242f31aca9cfdcc7c87f6a77eef6ced0c7967153400898e10dc264a196116d76" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "de/test-00000-of-00001.parquet", + "bytes": 4543635, + "sha256": "b58a0d5994e756dabef96eb804693be153953052276d3e0b7d57bb2f2798854a" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "ja/dev-00000-of-00001.parquet", + "bytes": 98924, + "sha256": "63c3fd6d20d016e7dc0df1d9b830ab5d93f9150450a24ec992a4662f2a80fefc" + }, + { + "repo": "CohereLabs/Global-MMLU", + "revision": "0e619dbeb34206cd48705a1a0ea7fb21cae09993", + "license": "apache-2.0", + "file": "ja/test-00000-of-00001.parquet", + "bytes": 4363616, + "sha256": "b56ad153180d4368757891796dfad0e1a76ebeb3f7b2500b1ac63789d4e4dfe7" + }, + { + "repo": "databricks/databricks-dolly-15k", + "revision": "bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a", + "license": "cc-by-sa-3.0", + "file": "README.md", + "bytes": 8199, + "sha256": "db06c60090a11b20e79d1933b2f316f781605f07761ff2ae2795d6523e3e960b" + }, + { + "repo": "databricks/databricks-dolly-15k", + "revision": "bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a", + "license": "cc-by-sa-3.0", + "file": "databricks-dolly-15k.jsonl", + "bytes": 13085339, + "sha256": "2df9083338b4abd6bceb5635764dab5d833b393b55759dffb0959b6fcbf794ec" + }, + { + "repo": "TIGER-Lab/MMLU-Pro", + "revision": "b189ec765aa7ed75c8acfea42df31fdae71f97be", + "file": "data/test-00000-of-00001.parquet", + "sha256": "0e24a191921c2f453518a537a8b2117bd137e7714d4ef1565e9ba06c1ecb9ad8", + "role": "historical exposure screening" + }, + { + "repo": "TIGER-Lab/MMLU-Pro", + "revision": "b189ec765aa7ed75c8acfea42df31fdae71f97be", + "file": "data/validation-00000-of-00001.parquet", + "sha256": "139423c23722e480c807ac4a191409a710cfce4eba744c1d641cf88e730e2078", + "role": "historical exposure screening" + }, + { + "repo": "LLM-Semantic-Router/category-classifier-supplement", + "revision": "c51aed0d3a83a548270e835187cee30615188e60", + "file": "data/train-00000-of-00001.parquet", + "sha256": "bf5c5e52f9c58f25d0e2a9bc7d23e401db20fde8fac6dece364a6820ed44b7d8", + "role": "historical exposure screening" + } +] diff --git a/reproducibility/src/training/model_classifier/sequence_repair/data/task-context-backgrounds.json b/reproducibility/src/training/model_classifier/sequence_repair/data/task-context-backgrounds.json new file mode 100644 index 0000000000000000000000000000000000000000..c6f28a9d068f7abf2e2d261a066151e2176cef91 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/data/task-context-backgrounds.json @@ -0,0 +1,14 @@ +{ + "train": { + "en": "Background notes from a furniture workshop. The apprentices set out pieces of plain timber on broad benches before the doors opened. Each surface had been sanded until the grain felt smooth beneath a folded cloth. A supervisor moved the unfinished frames away from the window and placed the small drawers beside their matching cabinets. The visitors examined the joints quietly and compared the texture of the wood with several unfinished samples.\n\nIn another corner, a display showed the steps involved in repairing an old chair. The loose seat had been removed so that the frame could rest on padded supports. Several tools lay in a shallow tray, with a sketch showing which part belonged on each side. During the break, the group arranged stools around the common table and discussed how to make room for a wider aisle. By the end of the afternoon the shavings had been swept into a basket, and the repaired pieces were left to dry. These notes describe background activity and contain no instruction for the requested task.", + "zh": "木工作坊的背景记录。学徒们在开门之前，把几块普通木料平放在宽大的工作台上。木板已经经过打磨，用折好的布轻轻擦过时，表面显得平整而光滑。负责人把尚未完成的框架从窗边移开，又将小抽屉放在配套的柜体旁边。来访者安静地观察榫卯结构，并把成品的纹理与几块未经上漆的样品相互比较。\n\n另一个角落展示了旧椅子的修补过程。松动的座板已经取下，椅架稳稳地放在软垫上。几件工具摆在浅盘里，旁边的草图标出了各个零件的位置。休息时，大家围着公共桌子摆好凳子，讨论怎样留出更宽的通道。到了下午结束，木屑被扫进篮子，修补好的部件则放在架子上晾干。这些文字只记录背景活动，不包含需要执行的任务。" + }, + "dev": { + "en": "Background account of a rehearsal space. Curtains softened the light entering the empty room while a group of volunteers unfolded the portable stage. Cushions had been stacked against the wall to leave the centre clear. The rehearsal involved moving slowly between several marked positions, pausing whenever a prop needed to be returned to its shelf. Everyone could see the doorway from the centre of the floor.\n\nAfter the first rehearsal, the team checked the arrangement of the folding screens. One screen had a plain cloth cover while another carried a pattern of leaves. The performers tried a quieter entrance and shifted the low platform slightly so that it would not block the passage. When the session ended, the curtains were tied back and the room was aired. The notes are reference background; the marked requested task appears separately.", + "zh": "排练室的背景记录。窗帘让空房间里的光线变得柔和，几位志愿者正在展开便携式舞台。靠垫整齐地堆在墙边，中央留出了活动空间。排练要求参与者在几个标记位置之间缓慢移动，需要把道具放回架子时便停下来。从地板中央能够清楚地看到门口。\n\n第一轮排练后，大家检查了折叠屏风的摆放方式。一扇屏风蒙着素色布面，另一扇则带着叶片图案。表演者尝试了更安静的入场方式，又把低矮的平台稍微挪开，避免挡住通道。活动结束时，窗帘被束起，房间里开始通风。这些笔记只是参考背景，明确标出的任务另列。" + }, + "test": { + "en": "Reference background from a shared courtyard. The residents had gathered empty flowerpots along the shaded side of a low wall. Some pots were tall and narrow, while others had a broad rim and a shallow base. A caretaker spread a woven mat beside the gate so that small tools could be laid out without getting lost in the grass. Fallen leaves were collected from the edges of the path, leaving the centre open for people carrying trays.\n\nLater, a light breeze lifted the edge of the mat. Two neighbours placed a smooth stone at each corner and moved the trays beneath an awning. They discussed the colour of the benches and the position of a hanging basket, then carried the remaining empty containers to the shed. The final entry described a quiet courtyard with a clear path and neatly stored tools. This passage supplies background description, while the explicitly marked request defines the task.", + "zh": "共享庭院的参考背景。居民们把空花盆放在矮墙背阴的一侧。有的花盆高而窄，有的边沿宽阔、底部较浅。管理员在门旁铺了一张编织垫，方便摆放小工具，免得工具落进草丛。大家清理了小路边缘的落叶，把中央留给端着托盘经过的人。\n\n后来，一阵轻风掀起了垫子的边角。两位邻居在每个角上压了一块光滑的石头，又把托盘移到遮雨棚下面。他们讨论了长椅的颜色和吊篮的位置，然后将剩下的空容器搬进储物间。最后一段记录写的是安静的庭院、畅通的小路以及整齐收好的工具。这里提供的是背景描述，任务由明确标出的请求决定。" + } +} diff --git a/reproducibility/src/training/model_classifier/sequence_repair/evaluate.py b/reproducibility/src/training/model_classifier/sequence_repair/evaluate.py new file mode 100644 index 0000000000000000000000000000000000000000..142b9862095d0a3c5a578316b60eb534125667ba --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/evaluate.py @@ -0,0 +1,147 @@ +"""Explicit streaming evaluation; never invoked on test data during training.""" + +# ruff: noqa: PLC0415 + +import argparse +import json +import time +from pathlib import Path + +from .data import classification_metrics, file_receipts, read_records +from .model import load_model + +LONG_BATCH_THRESHOLD = 2048 + + +def evaluate_records( + model, + tokenizer, + records, + label_to_id, + id_to_label, + max_length, + batch_size=16, + dtype="bfloat16", +): + import torch + + device = next(model.parameters()).device + if model.config.problem_type != "single_label_classification": + raise ValueError("Softmax evaluation requires a single-label classifier") + was_training = model.training + model.eval() + encoded = [ + tokenizer(row["text"], truncation=False, padding=False) for row in records + ] + lengths = [len(row["input_ids"]) for row in encoded] + if max(lengths) > max_length: + raise ValueError( + "Evaluation input exceeds the explicit budget; truncation is forbidden" + ) + order = sorted(range(len(records)), key=lengths.__getitem__) + results, timings, cursor = [None] * len(records), [], 0 + while cursor < len(order): + size = 1 if lengths[order[cursor]] > LONG_BATCH_THRESHOLD else batch_size + indices = order[cursor : cursor + size] + if any(lengths[index] > LONG_BATCH_THRESHOLD for index in indices): + indices = indices[:1] + cursor += len(indices) + batch = tokenizer.pad( + [encoded[index] for index in indices], padding=True, return_tensors="pt" + ) + batch = {key: value.to(device) for key, value in batch.items()} + if device.type == "cuda": + torch.cuda.synchronize() + start = time.perf_counter() + with torch.inference_mode(), torch.autocast( + device_type=device.type, + dtype=getattr(torch, dtype), + enabled=device.type == "cuda" and dtype != "float32", + ): + logits = model(**batch).logits.float() + if not bool(torch.isfinite(logits).all()): + raise ValueError("Non-finite evaluation logits") + probabilities = logits.softmax(-1).cpu().tolist() + if device.type == "cuda": + torch.cuda.synchronize() + timings.append((time.perf_counter() - start) * 1000) + for position, index in enumerate(indices): + scores = probabilities[position] + results[index] = { + "id": records[index]["id"], + "prediction": max(range(len(scores)), key=scores.__getitem__), + "probabilities": scores, + "tokens": lengths[index], + } + gold = [label_to_id[row["label"]] for row in records] + predicted = [row["prediction"] for row in results] + metrics = classification_metrics(gold, predicted, id_to_label) + metrics["breakdowns"] = {} + for field in ["source", "language", "length_bucket", "position"]: + metrics["breakdowns"][field] = {} + for group in sorted({str(row.get(field, "unspecified")) for row in records}): + indices = [ + i + for i, row in enumerate(records) + if str(row.get(field, "unspecified")) == group + ] + metrics["breakdowns"][field][group] = classification_metrics( + [gold[i] for i in indices], [predicted[i] for i in indices], id_to_label + ) + metrics["actual_tokens"] = {"min": min(lengths), "max": max(lengths)} + metrics["forward_batches_ms"] = timings + metrics["precision"] = { + "parameters": "float32", + "forward_autocast": ( + dtype if device.type == "cuda" and dtype != "float32" else None + ), + "device_type": device.type, + } + if was_training: + model.train() + return metrics, results + + +def main(): + import torch + + torch.set_num_threads(8) + parser = argparse.ArgumentParser() + parser.add_argument("--base", required=True) + parser.add_argument("--adapter") + parser.add_argument("--contract", required=True) + parser.add_argument("--data", nargs="+", required=True) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument("--max-length", type=int, default=32768) + parser.add_argument("--device", default="cuda") + parser.add_argument("--dtype", choices=["float32", "bfloat16"], default="bfloat16") + args = parser.parse_args() + if args.output.exists(): + raise ValueError("Refusing to overwrite evaluation evidence") + model, tokenizer, labels, ids = load_model(args.base, args.contract, args.adapter) + if args.max_length > model.config.max_position_embeddings: + raise ValueError("Evaluation budget exceeds checkpoint position capacity") + model.to(args.device) + rows = read_records(args.data, labels) + metrics, predictions = evaluate_records( + model, tokenizer, rows, labels, ids, args.max_length, dtype=args.dtype + ) + metrics["data_files"] = file_receipts(args.data) + args.output.parent.mkdir(parents=True, exist_ok=True) + args.output.write_text(json.dumps(metrics, indent=2) + "\n") + args.output.with_suffix(".predictions.jsonl").write_text( + "".join(json.dumps(row) + "\n" for row in predictions) + ) + print( + json.dumps( + { + key: metrics[key] + for key in ["rows", "accuracy", "macro_f1", "actual_tokens"] + } + ), + flush=True, + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/export.py b/reproducibility/src/training/model_classifier/sequence_repair/export.py new file mode 100644 index 0000000000000000000000000000000000000000..b4e57b59ce3569c034054c416761c92f18153d9f --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/export.py @@ -0,0 +1,106 @@ +"""Freeze a selected adapter as a standard HF merged sequence checkpoint.""" + +# ruff: noqa: PLC0415 + +import argparse +import hashlib +import json +from pathlib import Path + +from .model import load_model +from .runtime_mapping import RUNTIME_TASKS, write_runtime_mappings + + +def main(): + import torch + + parser = argparse.ArgumentParser() + parser.add_argument("--base", required=True) + parser.add_argument("--base-id", default="llm-semantic-router/mmbert-32k-yarn") + parser.add_argument("--base-revision", required=True) + parser.add_argument("--adapter", type=Path, required=True) + parser.add_argument("--contract", required=True) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument("--run-manifest", type=Path, required=True) + parser.add_argument("--runtime-task", choices=RUNTIME_TASKS, required=True) + args = parser.parse_args() + if args.output.exists() and any(args.output.iterdir()): + raise ValueError("Refusing to overwrite a frozen candidate") + run = json.loads(args.run_manifest.read_text()) + if run.get("test_used") is not False: + raise ValueError("Candidate provenance must explicitly exclude test selection") + if run.get("base_revision") != args.base_revision: + raise ValueError("Export base revision differs from the training receipt") + torch.set_num_threads(8) + model, tokenizer, labels, _ = load_model(args.base, args.contract, args.adapter) + model.eval() + tokens = tokenizer( + ["Numerical merge equivalence probe.", "用于核对合并数值的输入。"], + padding=True, + return_tensors="pt", + ) + with torch.inference_mode(): + before = model(**tokens).logits + merged = model.merge_and_unload(safe_merge=True) + after = merged(**tokens).logits + torch.testing.assert_close(before, after, atol=2e-4, rtol=2e-4) + if type(merged).__name__ != "ModernBertForSequenceClassification": + raise ValueError( + "Export must preserve the standard supported HF sequence architecture" + ) + if merged.config.label2id != labels: + raise ValueError("Merge changed label IDs") + merged.config._name_or_path = args.base_id + args.output.mkdir(parents=True, exist_ok=True) + merged.save_pretrained(args.output, safe_serialization=True) + tokenizer.save_pretrained(args.output) + mapping_files = write_runtime_mappings(args.output, labels, args.runtime_task) + receipt = { + "base_model": args.base_id, + "base_revision": args.base_revision, + "adapter_sha256": hashlib.sha256( + (args.adapter / "adapter_model.safetensors").read_bytes() + ).hexdigest(), + "run_manifest_sha256": hashlib.sha256( + args.run_manifest.read_bytes() + ).hexdigest(), + "parameter_count": sum(parameter.numel() for parameter in merged.parameters()), + "encoder_parameter_count": sum( + parameter.numel() for parameter in merged.model.parameters() + ), + "label2id": labels, + "runtime_task": args.runtime_task, + "runtime_mapping_files": mapping_files, + "task": merged.config.problem_type, + "classifier_pooling": merged.config.classifier_pooling, + "max_position_embeddings": merged.config.max_position_embeddings, + "position_capacity_is_quality_evidence": False, + "weight_dtype": "float32", + "test_used_for_selection": False, + "merge_probe_max_absolute_difference": float((before - after).abs().max()), + "files": { + path.name: hashlib.sha256(path.read_bytes()).hexdigest() + for path in sorted(args.output.iterdir()) + if path.is_file() + }, + } + (args.output / "candidate-lock.json").write_text( + json.dumps(receipt, indent=2) + "\n" + ) + print( + json.dumps( + { + key: receipt[key] + for key in [ + "parameter_count", + "merge_probe_max_absolute_difference", + "adapter_sha256", + ] + } + ), + flush=True, + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/fetch_sources.py b/reproducibility/src/training/model_classifier/sequence_repair/fetch_sources.py new file mode 100644 index 0000000000000000000000000000000000000000..e4b2045372cb5dbf6abd22be5622f74fcc641dbd --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/fetch_sources.py @@ -0,0 +1,48 @@ +"""Fetch only the pinned public dataset files used by the repair recipes.""" + +import argparse +import hashlib +import json +from pathlib import Path + + +def main(): + from huggingface_hub import hf_hub_download # noqa: PLC0415 + + parser = argparse.ArgumentParser() + parser.add_argument( + "--manifest", + type=Path, + default=Path(__file__).parent / "data/source-files.json", + ) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + receipts = [] + for item in json.loads(args.manifest.read_text()): + directory = args.output / item["repo"].replace("/", "--") + path = Path( + hf_hub_download( + repo_id=item["repo"], + repo_type="dataset", + filename=item["file"], + revision=item["revision"], + local_dir=directory, + ) + ) + digest = hashlib.sha256(path.read_bytes()).hexdigest() + if digest != item["sha256"]: + raise ValueError( + f"Pinned dataset file hash differs: {item['repo']}/{item['file']}" + ) + receipts.append({**item, "verified_sha256": digest}) + print( + json.dumps({"repo": item["repo"], "file": item["file"], "verified": True}), + flush=True, + ) + (args.output / "verified-source-files.json").write_text( + json.dumps(receipts, indent=2) + "\n" + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/hydrate_annotations.py b/reproducibility/src/training/model_classifier/sequence_repair/hydrate_annotations.py new file mode 100644 index 0000000000000000000000000000000000000000..4cc2d7c6d40ebd9403b357081fe4f918521b5d35 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/hydrate_annotations.py @@ -0,0 +1,95 @@ +"""Reconstruct reviewed requests from pinned source data and text-free label sidecars.""" + +# ruff: noqa: PLC0415 +import argparse +import hashlib +import json +from pathlib import Path + + +def sha256_text(text): + return hashlib.sha256(text.encode()).hexdigest() + + +def hydrate_records(sidecar, sources): + repository = sidecar["source"]["repository"] + result = [] + if repository == "CohereLabs/aya_dataset": + lookup = {} + for row in sources: + if not row.get("user_id"): + continue + key = ( + sha256_text(row["inputs"]), + row["language_code"], + "aya-author-" + sha256_text(row["user_id"]), + ) + lookup[key] = row["inputs"] + for row in sidecar["records"]: + key = (row["source_text_sha256"], row["language"], row["group_id"]) + if key not in lookup: + raise ValueError( + "Reviewed Aya request or author differs from the pinned source" + ) + result.append({**row, "text": lookup[key]}) + elif repository == "databricks/databricks-dolly-15k": + for row in sidecar["records"]: + source = sources[row["source_row"]] + instruction, context = ( + source["instruction"].strip(), + source["context"].strip(), + ) + text = instruction + ("\n\n" + context if context else "") + if sha256_text(text) != row["source_text_sha256"]: + raise ValueError( + "Reviewed Dolly request differs from the pinned source" + ) + result.append( + {**row, "text": text, "instruction": instruction, "context": context} + ) + else: + raise ValueError("Unknown reviewed source") + return result + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument("--sidecar", type=Path, required=True) + parser.add_argument("--source", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + if args.output.exists(): + raise ValueError("Refusing to overwrite reviewed annotations") + sidecar = json.loads(args.sidecar.read_text()) + if ( + hashlib.sha256(args.source.read_bytes()).hexdigest() + != sidecar["source"]["file_sha256"] + ): + raise ValueError("Source file hash differs from the reviewed revision") + if sidecar["source"]["repository"] == "CohereLabs/aya_dataset": + from pyarrow import parquet + + sources = parquet.read_table( + args.source, columns=["inputs", "language_code", "user_id"] + ).to_pylist() + else: + sources = [ + json.loads(line) + for line in args.source.read_text().split("\n") + if line.strip() + ] + rows = hydrate_records(sidecar, sources) + content = "".join( + json.dumps(row, ensure_ascii=False, sort_keys=True) + "\n" for row in rows + ) + args.output.parent.mkdir(parents=True, exist_ok=True) + args.output.write_text(content) + print( + json.dumps( + {"rows": len(rows), "sha256": hashlib.sha256(content.encode()).hexdigest()} + ) + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/initialize.py b/reproducibility/src/training/model_classifier/sequence_repair/initialize.py new file mode 100644 index 0000000000000000000000000000000000000000..b011ec55f3f8d6138a4afdeee88c729558f36501 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/initialize.py @@ -0,0 +1,73 @@ +"""Initialize a fresh task adapter and trainable head from a fixed encoder base.""" + +# ruff: noqa: PLC0415 +import argparse +import hashlib +import json +from pathlib import Path + +from .model import load_model, save_adapter + + +def main(): + import torch + from peft import LoraConfig, TaskType, get_peft_model + + parser = argparse.ArgumentParser() + parser.add_argument("--base", required=True) + parser.add_argument("--base-id", default="llm-semantic-router/mmbert-32k-yarn") + parser.add_argument("--base-revision", required=True) + parser.add_argument("--contract", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument("--seed", type=int, default=20260913) + parser.add_argument("--rank", type=int, default=32) + parser.add_argument("--alpha", type=int, default=64) + parser.add_argument("--dropout", type=float, default=0.05) + args = parser.parse_args() + if args.output.exists(): + raise ValueError("Refusing to overwrite an initialized adapter") + if min(args.rank, args.alpha) <= 0 or not 0 <= args.dropout < 1: + raise ValueError("Invalid LoRA dimensions or dropout") + torch.set_num_threads(8) + torch.manual_seed(args.seed) + model, tokenizer, _labels, _ids = load_model(args.base, args.contract) + model = get_peft_model( + model, + LoraConfig( + task_type=TaskType.SEQ_CLS, + r=args.rank, + lora_alpha=args.alpha, + lora_dropout=args.dropout, + target_modules=["attn.Wqkv", "attn.Wo", "mlp.Wi", "mlp.Wo"], + modules_to_save=["head", "classifier"], + bias="none", + revision=args.base_revision, + ), + ) + save_adapter(model, tokenizer, args.output, args.base_id, args.base_revision) + (args.output / "contract.json").write_bytes(args.contract.read_bytes()) + receipt = { + "base_model": args.base_id, + "base_revision": args.base_revision, + "seed": args.seed, + "rank": args.rank, + "alpha": args.alpha, + "dropout": args.dropout, + "classifier_pooling": model.config.classifier_pooling, + "adapter_sha256": hashlib.sha256( + (args.output / "adapter_model.safetensors").read_bytes() + ).hexdigest(), + "trainable_parameters": sum( + parameter.numel() + for parameter in model.parameters() + if parameter.requires_grad + ), + } + (args.output / "initialization.json").write_text( + json.dumps(receipt, indent=2) + "\n" + ) + print(json.dumps(receipt)) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/model.py b/reproducibility/src/training/model_classifier/sequence_repair/model.py new file mode 100644 index 0000000000000000000000000000000000000000..5c9a2ea710fa111110908257a8b0a2983eac2bda --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/model.py @@ -0,0 +1,71 @@ +"""HF standard sequence checkpoints with an unchanged label contract.""" + +# Heavy imports stay optional for data-contract tests. +# ruff: noqa: PLC0415 + +import json +from pathlib import Path + +from .data import load_contract + + +def configuration_overrides(contract): + """Validate optional standard HF task settings without guessing a pooling mode.""" + content = json.loads(Path(contract).read_text()) + problem_type = content.get("problem_type", "single_label_classification") + if problem_type not in { + "single_label_classification", + "multi_label_classification", + }: + raise ValueError("Unsupported classification problem type") + overrides = {"problem_type": problem_type} + if "classifier_pooling" in content: + pooling = content["classifier_pooling"] + if pooling not in {"mean", "cls"}: + raise ValueError("Sequence pooling must be explicitly mean or cls") + overrides["classifier_pooling"] = pooling + return overrides + + +def load_model(base, contract, adapter=None, trainable=False): + import torch + from peft import PeftModel + from transformers import AutoModelForSequenceClassification, AutoTokenizer + + label_to_id, id_to_label = load_contract(contract) + overrides = configuration_overrides(contract) + tokenizer = AutoTokenizer.from_pretrained(base) + model = AutoModelForSequenceClassification.from_pretrained( + base, + num_labels=len(label_to_id), + label2id=label_to_id, + id2label=id_to_label, + torch_dtype=torch.float32, + attn_implementation="sdpa", + reference_compile=False, + **overrides, + ) + if adapter: + model = PeftModel.from_pretrained(model, adapter, is_trainable=trainable) + elif trainable: + raise ValueError("Continuation requires an explicit initial adapter") + if model.config.label2id != label_to_id: + raise ValueError("The loaded classifier changed the label contract") + return model, tokenizer, label_to_id, id_to_label + + +def save_adapter(model, tokenizer, destination, base_id, base_revision): + import json + + destination.mkdir(parents=True, exist_ok=True) + model.peft_config["default"].base_model_name_or_path = base_id + model.peft_config["default"].revision = base_revision + model.save_pretrained(destination) + tokenizer.save_pretrained(destination) + (destination / "label_mapping.json").write_text( + json.dumps( + {"label2id": model.config.label2id, "id2label": model.config.id2label}, + indent=2, + ) + + "\n" + ) diff --git a/reproducibility/src/training/model_classifier/sequence_repair/prepare_context.py b/reproducibility/src/training/model_classifier/sequence_repair/prepare_context.py new file mode 100644 index 0000000000000000000000000000000000000000..9c34a1e399269d6efb813aae1ebb686611905e5f --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/prepare_context.py @@ -0,0 +1,138 @@ +"""Frozen context curriculum and held-out stress variants from task source groups.""" + +# Chinese punctuation is intentional in language-specific task markers. +# ruff: noqa: PLC0415, RUF001 +import argparse +import hashlib +import json +from collections import defaultdict +from pathlib import Path + +from .context import insert_payload +from .corpus import write_corpus +from .data import read_records + +LANGUAGE_ALIASES = { + "eng": "en", + "zho": "zh", + "spa": "es", + "fra": "fr", + "deu": "de", + "jpn": "ja", +} + + +def choose_payloads(rows, languages, per_label): + pools = defaultdict(list) + for row in rows: + language = LANGUAGE_ALIASES.get(row.get("language"), row.get("language")) + if language in languages: + pools[(language, row["label"])].append(row) + result = [] + for (language, _label), pool in sorted(pools.items()): + seen = set() + for row in sorted( + pool, key=lambda item: hashlib.sha256(item["id"].encode()).hexdigest() + ): + if row["group_id"] in seen: + continue + seen.add(row["group_id"]) + result.append({**row, "language": language}) + if len(seen) >= per_label: + break + return result + + +def main(): + from transformers import AutoTokenizer + + parser = argparse.ArgumentParser() + parser.add_argument("--corpus", type=Path, required=True) + parser.add_argument("--tokenizer", required=True) + parser.add_argument("--backgrounds", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument("--train-per-label", type=int, default=3) + parser.add_argument("--eval-per-label", type=int, default=1) + parser.add_argument( + "--budgets", type=int, nargs="+", default=[4096, 8192, 16384, 32768] + ) + parser.add_argument("--languages", nargs="+", default=["en", "zh"]) + parser.add_argument("--payload-sources", nargs="+") + args = parser.parse_args() + if min(args.train_per_label, args.eval_per_label, *args.budgets) <= 0: + raise ValueError("Context budgets and sample counts must be positive") + backgrounds = json.loads(args.backgrounds.read_text()) + if len({text for group in backgrounds.values() for text in group.values()}) != sum( + len(group) for group in backgrounds.values() + ): + raise ValueError("Background text must differ across languages and partitions") + tokenizer = AutoTokenizer.from_pretrained(args.tokenizer) + contract = json.loads((args.corpus / "contract.json").read_text()) + + def count(text): + return len(tokenizer(text, truncation=False, padding=False)["input_ids"]) + + splits = {} + for split in ["train", "dev", "test"]: + rows = read_records([args.corpus / f"{split}.jsonl"], contract["label2id"]) + if args.payload_sources: + rows = [row for row in rows if row.get("source") in args.payload_sources] + chosen = choose_payloads( + rows, + args.languages, + args.train_per_label if split == "train" else args.eval_per_label, + ) + if not chosen: + raise ValueError(f"No eligible {split} source groups") + output = [] + for row in chosen: + language = row["language"] + prefix = "Requested task:\n" if language == "en" else "需要完成的任务：\n" + payload = prefix + row["text"] + for budget in args.budgets: + for position in ["head", "middle", "tail"]: + built = insert_payload( + payload, backgrounds[split][language], position, budget, count + ) + output.append( + { + **row, + "id": f"context:{row['id']}:{budget}:{position}", + "text": built["text"], + "source": "authored-context-on-frozen-source", + "parent_source": row["source"], + "parent_id": row["id"], + "length_bucket": str(budget), + "position": position, + "actual_tokens": built["actual_tokens"], + "payload_start": built["payload_start"], + "payload_end": built["payload_end"], + } + ) + splits[split] = output + print( + json.dumps( + {"split": split, "payloads": len(chosen), "variants": len(output)} + ), + flush=True, + ) + provenance = { + "parent_manifest_sha256": hashlib.sha256( + (args.corpus / "manifest.json").read_bytes() + ).hexdigest(), + "backgrounds_sha256": hashlib.sha256(args.backgrounds.read_bytes()).hexdigest(), + "tokenizer_sha256": hashlib.sha256( + (Path(args.tokenizer) / "tokenizer.json").read_bytes() + ).hexdigest(), + "train_payloads_per_label_language": args.train_per_label, + "eval_payloads_per_label_language": args.eval_per_label, + "budgets": args.budgets, + "languages": args.languages, + "scope": "Authored repeated multi-paragraph context around frozen source requests, not natural long-document accuracy. Label is the explicitly marked requested task. Paired length/position variants retain the parent source group and partition.", + "test_used": False, + } + print(json.dumps(write_corpus(args.output, splits, contract, provenance), indent=2)) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/runtime_mapping.py b/reproducibility/src/training/model_classifier/sequence_repair/runtime_mapping.py new file mode 100644 index 0000000000000000000000000000000000000000..4b361c198bff6ca472c5711c3bda6effa342737f --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/runtime_mapping.py @@ -0,0 +1,58 @@ +"""Serialize classifier label sidecars consumed by the router's Go loaders.""" + +import json +from pathlib import Path + +RUNTIME_TASKS = ( + "domain", + "fact-check", + "modality", + "feedback", + "prompt-guard", + "safety", + "hazard", +) + + +def write_runtime_mappings(destination, labels, task): + """Write both HF and router layouts from one contiguous label bijection.""" + if task not in RUNTIME_TASKS: + raise ValueError("Unknown runtime classifier task") + if not labels or set(labels.values()) != set(range(len(labels))): + raise ValueError("Runtime label IDs must be a contiguous bijection") + if task == "modality" and labels != {"AR": 0, "DIFFUSION": 1, "BOTH": 2}: + raise ValueError("Modality must preserve the native AR/DIFFUSION/BOTH IDs") + if task == "fact-check" and set(labels) != { + "NO_FACT_CHECK_NEEDED", + "FACT_CHECK_NEEDED", + }: + raise ValueError("FactCheck requires the router's two semantic labels") + reverse = {str(index): label for label, index in labels.items()} + generic = { + "label2id": labels, + "id2label": reverse, + "label_to_idx": labels, + "idx_to_label": reverse, + } + files = {"label_mapping.json": generic} + if task == "domain": + files["category_mapping.json"] = { + "category_to_idx": labels, + "idx_to_category": reverse, + } + sidecars = { + "fact-check": "fact_check_mapping.json", + "modality": "modality_mapping.json", + "prompt-guard": "jailbreak_type_mapping.json", + } + if task in sidecars: + files[sidecars[task]] = { + "label_to_idx": labels, + "idx_to_label": reverse, + } + for name, data in files.items(): + path = Path(destination) / name + if path.exists(): + raise ValueError("Refusing to overwrite an existing runtime mapping") + path.write_text(json.dumps(data, indent=2) + "\n") + return sorted(files) diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_annotations.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_annotations.py new file mode 100644 index 0000000000000000000000000000000000000000..622bb7a771d0e68f703b68e0c2b3ebdeaa93b3d5 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_annotations.py @@ -0,0 +1,58 @@ +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) +from src.training.model_classifier.sequence_repair.hydrate_annotations import ( + hydrate_records, + sha256_text, +) + + +class AnnotationTests(unittest.TestCase): + def test_aya_author_and_text_both_identify_the_reviewed_request(self): + source = { + "inputs": "读这段文本\u2028再总结。", + "language_code": "zho", + "user_id": "public-contributor", + } + sidecar = { + "source": {"repository": "CohereLabs/aya_dataset"}, + "records": [ + { + "id": "review", + "source_text_sha256": sha256_text(source["inputs"]), + "language": "zho", + "group_id": "aya-author-" + sha256_text(source["user_id"]), + "label": "NO_FACT_CHECK_NEEDED", + } + ], + } + result = hydrate_records(sidecar, [source]) + self.assertEqual(result[0]["text"], source["inputs"]) + with self.assertRaises(ValueError): + hydrate_records(sidecar, [{**source, "user_id": "another-contributor"}]) + + def test_dolly_context_is_reconstructed_and_verified_not_discarded(self): + source = { + "instruction": " Extract the year. ", + "context": " The club opened in 1910. ", + } + expected = "Extract the year.\n\nThe club opened in 1910." + sidecar = { + "source": {"repository": "databricks/databricks-dolly-15k"}, + "records": [ + { + "source_row": 0, + "source_text_sha256": sha256_text(expected), + "label": "NO_FACT_CHECK_NEEDED", + } + ], + } + self.assertEqual(hydrate_records(sidecar, [source])[0]["text"], expected) + with self.assertRaises(ValueError): + hydrate_records(sidecar, [{**source, "context": ""}]) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_context.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_context.py new file mode 100644 index 0000000000000000000000000000000000000000..e190f26b8322bfcf7a5b2503e6987e2d2d4ae497 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_context.py @@ -0,0 +1,76 @@ +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) +from src.training.model_classifier.sequence_repair.context import insert_payload +from src.training.model_classifier.sequence_repair.prepare_context import ( + choose_payloads, +) + + +class ContextTests(unittest.TestCase): + def test_estimated_search_matches_complete_paragraph_brute_force(self): + payload, background = "task", "a b c" + for budget in [5, 8, 31, 127, 513]: + for position in ["head", "middle", "tail"]: + # Boundary variation makes a linear estimate imperfect. + def count(text): + return len(text.split()) + 2 + text.count("\n") // 3 + + expected = 0 + for repetitions in range(budget + 1): + left = {"head": 0, "middle": repetitions // 2, "tail": repetitions}[ + position + ] + parts = ( + [background] * left + + [payload] + + [background] * (repetitions - left) + ) + if count("\n".join(parts)) <= budget: + expected = repetitions + result = insert_payload( + payload, background, position, budget, count, fill_to_budget=False + ) + self.assertEqual(result["text"].count(background), expected) + + def test_payload_selection_preserves_distinct_source_groups_and_language(self): + rows = [ + {"id": "first", "group_id": "same", "label": "yes", "language": "eng"}, + {"id": "variant", "group_id": "same", "label": "yes", "language": "en"}, + {"id": "other", "group_id": "other", "label": "yes", "language": "en"}, + {"id": "outside", "group_id": "fr", "label": "yes", "language": "fra"}, + ] + selected = choose_payloads(rows, ["en"], 2) + self.assertEqual({row["group_id"] for row in selected}, {"same", "other"}) + self.assertEqual({row["language"] for row in selected}, {"en"}) + + def test_payload_survives_every_position_at_exact_attended_budget(self): + def count(text): + return len(text.split()) + 2 + + for position in ["head", "middle", "tail"]: + result = insert_payload( + "用户名 alice@example.test", + "This is ordinary background.", + position, + 128, + count, + ) + self.assertEqual(result["actual_tokens"], 128) + self.assertEqual( + result["text"][result["payload_start"] : result["payload_end"]], + "用户名 alice@example.test", + ) + self.assertEqual(result["text"].count("alice@example.test"), 1) + + def test_oversized_payload_is_rejected_without_truncation(self): + with self.assertRaises(ValueError): + insert_payload( + "one two three", "background", "tail", 2, lambda text: len(text.split()) + ) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_contracts.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_contracts.py new file mode 100644 index 0000000000000000000000000000000000000000..630dc071e189937f3379b73dcda86a3f3d338a68 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_contracts.py @@ -0,0 +1,195 @@ +import importlib.util +import json +import random +import sys +import tempfile +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) +from src.training.model_classifier.sequence_repair.data import ( + assert_disjoint, + classification_metrics, + load_contract, + read_records, +) +from src.training.model_classifier.sequence_repair.model import configuration_overrides +from src.training.model_classifier.sequence_repair.train import ( + length_sampling_pools, + microbatch_loss, + sample_length_balanced, + sample_source_balanced, + selection_score, + source_sampling_pools, +) + + +class ContractTests(unittest.TestCase): + def test_explicit_pooling_preserves_legacy_default_and_rejects_unknown_modes(self): + with tempfile.TemporaryDirectory() as directory: + path = Path(directory) / "contract.json" + path.write_text("{}") + self.assertNotIn("classifier_pooling", configuration_overrides(path)) + for pooling in ["mean", "cls"]: + path.write_text(json.dumps({"classifier_pooling": pooling})) + self.assertEqual( + configuration_overrides(path)["classifier_pooling"], pooling + ) + for pooling in [None, "last", "CLS"]: + path.write_text(json.dumps({"classifier_pooling": pooling})) + with self.assertRaises(ValueError): + configuration_overrides(path) + + def test_small_reviewed_source_remains_visible_with_length_and_label_balance(self): + rows = [{"source": "weak", "length_bucket": "short", "label": "a"}] * 1000 + [ + {"source": "reviewed", "length_bucket": length, "label": label} + for length in ["short", 32768] + for label in ["a", "b"] + ] + pools = source_sampling_pools(rows, balance_lengths=True) + rng = random.Random(17) + counts = dict.fromkeys(range(1000, 1004), 0) + for _ in range(4000): + index = sample_source_balanced(pools, rng, True) + if index in counts: + counts[index] += 1 + for count in counts.values(): + self.assertGreater(count, 400) + self.assertLess(count, 600) + with self.assertRaises(ValueError): + source_sampling_pools([{"label": "a"}], False) + + def test_source_selection_does_not_hide_a_reviewed_source_regression(self): + metrics = { + "macro_f1": 0.93, + "breakdowns": { + "source": {"weak": {"macro_f1": 0.99}, "reviewed": {"macro_f1": 0.5}} + }, + } + self.assertAlmostEqual(selection_score(metrics, "source-macro-f1"), 0.745) + + def test_length_sampling_keeps_a_rare_long_bucket_visible(self): + rows = [{"length_bucket": "short", "label": "a"}] * 1000 + [ + {"length_bucket": 32768, "label": "a"} + ] + pools = length_sampling_pools(rows) + rng = random.Random(42) + long_samples = sum( + sample_length_balanced(pools, rng, True) == len(rows) - 1 + for _ in range(1000) + ) + self.assertGreater(long_samples, 400) + self.assertLess(long_samples, 600) + + def test_unicode_separator_inside_json_string_is_not_a_record_boundary(self): + with tempfile.TemporaryDirectory() as directory: + path = Path(directory) / "data.jsonl" + row = { + "id": "unicode", + "text": "one\u2028two\u2029three\u0085four", + "group_id": "unicode", + "label": "x", + } + path.write_text(json.dumps(row, ensure_ascii=False) + "\n") + self.assertEqual(read_records([path], {"x": 0}), [row]) + + def test_label_ids_are_bijective_not_resorted(self): + with tempfile.TemporaryDirectory() as directory: + path = Path(directory) / "config.json" + path.write_text( + json.dumps( + {"id2label": {"0": "z", "1": "a"}, "label2id": {"z": 0, "a": 1}} + ) + ) + self.assertEqual(load_contract(path)[0], {"z": 0, "a": 1}) + path.write_text( + json.dumps( + {"id2label": {"0": "z", "1": "a"}, "label2id": {"a": 0, "z": 1}} + ) + ) + with self.assertRaises(ValueError): + load_contract(path) + + def test_unicode_normalized_text_and_translation_groups_cannot_leak(self): + train = [{"id": "a", "text": "\uff21 cat", "group_id": "question-1"}] + for dev in [ + [{"id": "b", "text": "a cat", "group_id": "other"}], + [{"id": "b", "text": "一只猫", "group_id": "question-1"}], + ]: + with self.assertRaises(ValueError): + assert_disjoint(train, dev) + assert_disjoint( + train, [{"id": "c", "text": "new question", "group_id": "question-2"}] + ) + + def test_conflicting_annotations_are_rejected(self): + with tempfile.TemporaryDirectory() as directory: + path = Path(directory) / "data.jsonl" + rows = [ + {"id": "a", "text": "Hello", "group_id": "g1", "label": "x"}, + {"id": "b", "text": "hello", "group_id": "g2", "label": "y"}, + ] + path.write_text("\n".join(map(json.dumps, rows))) + with self.assertRaises(ValueError): + read_records([path], {"x": 0, "y": 1}) + + def test_macro_f1_exposes_failure_on_minority_label(self): + metrics = classification_metrics( + [0] * 9 + [1], [0] * 10, {0: "majority", 1: "minority"} + ) + self.assertAlmostEqual(metrics["accuracy"], 0.9) + self.assertEqual(metrics["per_label"]["minority"]["recall"], 0) + self.assertLess(metrics["macro_f1"], 0.5) + + def test_length_selection_cannot_be_dominated_by_dense_bucket(self): + metrics = { + "macro_f1": 0.9, + "breakdowns": { + "length_bucket": {"short": {"macro_f1": 0.9}, "long": {"macro_f1": 0.1}} + }, + } + self.assertAlmostEqual(selection_score(metrics, "length-macro-f1"), 0.5) + + def test_present_source_selection_does_not_penalize_unobserved_labels(self): + labels = {i: str(i) for i in range(5)} + metrics = { + "breakdowns": { + "source": { + "legacy_four": classification_metrics( + [0, 1, 2, 3], [0, 1, 2, 3], labels + ), + "new_negative": classification_metrics([4], [4], labels), + } + } + } + self.assertAlmostEqual(selection_score(metrics, "source-macro-f1"), 0.5) + self.assertAlmostEqual(selection_score(metrics, "source-present-macro-f1"), 1.0) + metrics["breakdowns"]["source"]["new_negative"] = classification_metrics( + [4], [0], labels + ) + self.assertAlmostEqual(selection_score(metrics, "source-present-macro-f1"), 0.5) + + @unittest.skipUnless( + importlib.util.find_spec("torch"), + "Optional real tensor gradient test requires Torch", + ) + def test_accumulated_mean_cross_entropy_matches_full_batch_gradient(self): + import torch # noqa: PLC0415 - optional tensor test + + torch.manual_seed(7) + inputs = torch.randn(16, 5, dtype=torch.float64) + labels = torch.arange(16) % 3 + full = torch.nn.Linear(5, 3, dtype=torch.float64) + micro = torch.nn.Linear(5, 3, dtype=torch.float64) + micro.load_state_dict(full.state_dict()) + microbatch_loss(full(inputs), labels, 1).backward() + for start in range(0, 16, 2): + microbatch_loss( + micro(inputs[start : start + 2]), labels[start : start + 2], 8 + ).backward() + for expected, actual in zip(full.parameters(), micro.parameters(), strict=True): + torch.testing.assert_close(expected.grad, actual.grad, rtol=1e-6, atol=1e-7) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_recipe_sources.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_recipe_sources.py new file mode 100644 index 0000000000000000000000000000000000000000..d3bd6e2d6d37ece226504041bc7cd99cafa246c6 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_recipe_sources.py @@ -0,0 +1,65 @@ +import hashlib +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) +from src.training.model_classifier.domain_classifier.prepare_data import ( + canonical_groups, + question_with_choices, +) +from src.training.model_classifier.modality_routing_classifier.extend_contracts import ( + recover_caption, +) +from src.training.model_classifier.modality_routing_classifier.prepare_vela_data import ( + TEMPLATES, +) +from src.training.model_classifier.sequence_repair.data import normalized_text + + +class SourceContractTests(unittest.TestCase): + def test_translated_duplicate_groups_are_unioned_transitively(self): + rows = [ + {"text": "First question", "group_id": "a", "label": "biology"}, + {"text": "FIRST QUESTION", "group_id": "b", "label": "biology"}, + {"text": "Otra pregunta", "group_id": "b", "label": "biology"}, + {"text": "otra pregunta", "group_id": "c", "label": "biology"}, + ] + retained, excluded = canonical_groups(rows) + self.assertEqual(excluded, 0) + self.assertEqual({row["group_id"] for row in retained}, {"a"}) + retained, excluded = canonical_groups( + [*rows, {"text": "Otra pregunta", "group_id": "d", "label": "history"}] + ) + self.assertEqual((retained, excluded), ([], 5)) + + def test_domain_question_contains_choices_but_not_answer_key(self): + row = { + "question": "Which element?", + "option_a": "iron", + "option_b": "wood", + "option_c": "paper", + "option_d": "glass", + "answer": "secret-answer-key", + } + text = question_with_choices(row) + self.assertIn("A. iron", text) + self.assertNotIn(row["answer"], text) + + def test_caption_recovery_keeps_embedded_colons_and_rejects_changed_source(self): + caption = "A quiet stage: blue curtains, paper lanterns, soft light" + row = { + "language": "en", + "template_family": "modality-train-1", + "text": TEMPLATES["en"]["AR"][1].format(caption), + "caption_sha256": hashlib.sha256( + normalized_text(caption).encode() + ).hexdigest(), + } + self.assertEqual(recover_caption(row), caption) + with self.assertRaises(ValueError): + recover_caption({**row, "text": row["text"] + " changed"}) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_runtime_mapping.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_runtime_mapping.py new file mode 100644 index 0000000000000000000000000000000000000000..dfdb812acc39e5b9948e6ee0aba54615a1a2dfdf --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_runtime_mapping.py @@ -0,0 +1,69 @@ +"""Final model layouts must agree with both HF and router label consumers.""" + +import json +import tempfile +import unittest +from pathlib import Path + +from src.training.model_classifier.sequence_repair.runtime_mapping import ( + write_runtime_mappings, +) + + +class RuntimeMappingTests(unittest.TestCase): + def test_domain_uses_category_loader_schema_and_same_ids(self): + with tempfile.TemporaryDirectory() as directory: + labels = {"biology": 0, "other": 1} + files = write_runtime_mappings(directory, labels, "domain") + self.assertEqual(files, ["category_mapping.json", "label_mapping.json"]) + category = json.loads( + (Path(directory) / "category_mapping.json").read_text() + ) + generic = json.loads((Path(directory) / "label_mapping.json").read_text()) + self.assertEqual(category["category_to_idx"], generic["label2id"]) + self.assertEqual(category["idx_to_category"], generic["id2label"]) + + def test_factcheck_polarity_is_preserved_in_both_directions(self): + with tempfile.TemporaryDirectory() as directory: + labels = {"FACT_CHECK_NEEDED": 0, "NO_FACT_CHECK_NEEDED": 1} + write_runtime_mappings(directory, labels, "fact-check") + mapping = json.loads( + (Path(directory) / "fact_check_mapping.json").read_text() + ) + self.assertEqual(mapping["label_to_idx"], labels) + for label, index in labels.items(): + self.assertEqual(mapping["idx_to_label"][str(index)], label) + + def test_modality_cannot_reorder_native_label_indices(self): + with tempfile.TemporaryDirectory() as directory: + with self.assertRaisesRegex(ValueError, "native"): + write_runtime_mappings( + directory, {"AR": 1, "DIFFUSION": 0, "BOTH": 2}, "modality" + ) + self.assertEqual(list(Path(directory).iterdir()), []) + + def test_prompt_guard_and_dynamic_labels_use_router_layout(self): + for task in ("prompt-guard", "feedback", "safety", "hazard"): + with self.subTest(task=task), tempfile.TemporaryDirectory() as directory: + labels = {"safe": 0, "unsafe": 1} + files = write_runtime_mappings(directory, labels, task) + name = ( + "jailbreak_type_mapping.json" + if task == "prompt-guard" + else "label_mapping.json" + ) + self.assertIn(name, files) + mapping = json.loads((Path(directory) / name).read_text()) + self.assertEqual(mapping["label_to_idx"], labels) + self.assertEqual(mapping["idx_to_label"], {"0": "safe", "1": "unsafe"}) + + def test_sparse_or_duplicate_ids_fail_before_writing(self): + for labels in ({"a": 0, "b": 0}, {"a": 0, "b": 2}): + with tempfile.TemporaryDirectory() as directory: + with self.assertRaisesRegex(ValueError, "bijection"): + write_runtime_mappings(directory, labels, "domain") + self.assertEqual(list(Path(directory).iterdir()), []) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_sampling.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_sampling.py new file mode 100644 index 0000000000000000000000000000000000000000..ef1c0232a979c35b3b80e77ecb5245c32da600fc --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_sampling.py @@ -0,0 +1,82 @@ +import json +import random +import sys +import unittest +from collections import Counter +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) + +from src.training.model_classifier.sequence_repair.train import ( + sample_length_balanced, + sample_source_balanced, + sampling_exposure, + source_sampling_pools, + source_sampling_weights, +) + + +class SourceSamplingTests(unittest.TestCase): + def setUp(self): + self.rows = [ + {"source": "natural", "label": "a", "length_bucket": 10}, + {"source": "natural", "label": "b", "length_bucket": 100}, + {"source": "authored", "label": "a", "length_bucket": 10}, + ] + + def test_missing_weights_preserve_legacy_rng_sequence(self): + pools = source_sampling_pools(self.rows, True) + actual, legacy = random.Random(81), random.Random(81) + weights = source_sampling_weights(self.rows, None) + for _ in range(500): + self.assertEqual( + sample_source_balanced(pools, actual, True, weights), + sample_length_balanced(legacy.choice(pools), legacy, True), + ) + self.assertEqual(actual.getstate(), legacy.getstate()) + + def test_weight_order_uses_source_occurrence_and_controls_frequency(self): + weights = source_sampling_weights(self.rows, '{"authored": 1, "natural": 9}') + self.assertEqual(weights, [0.9, 0.1]) + pools = source_sampling_pools(self.rows, True) + rng = random.Random(81) + drawn = Counter( + sample_source_balanced(pools, rng, True, weights) for _ in range(10000) + ) + self.assertGreater(drawn[2], 900) + self.assertLess(drawn[2], 1100) + self.assertGreater(drawn[0], 4000) + self.assertGreater(drawn[1], 4000) + exposure = sampling_exposure(self.rows, drawn) + self.assertEqual(exposure["natural"]["unique_rows"], 2) + self.assertEqual(exposure["natural"]["draws"], drawn[0] + drawn[1]) + self.assertEqual(exposure["natural"]["label_draws"]["b"], drawn[1]) + + def test_invalid_or_incomplete_source_weights_are_rejected(self): + for specification in [ + {}, + [], + {"natural": 1}, + {"natural": 1, "authored": 1, "unknown": 1}, + *( + {"natural": value, "authored": 1} + for value in [True, "1", 0, -1, float("nan"), float("inf")] + ), + {"natural": 1e308, "authored": 1e308}, + ]: + with self.subTest(specification=specification), self.assertRaises( + ValueError + ): + source_sampling_weights(self.rows, json.dumps(specification)) + with self.assertRaises(ValueError): + source_sampling_weights([{"label": "a"}], '{"natural": 1}') + + def test_unique_coverage_keeps_unseen_rows_visible(self): + exposure = sampling_exposure(self.rows, Counter({0: 100, 2: 5})) + self.assertEqual(exposure["natural"]["rows"], 2) + self.assertEqual(exposure["natural"]["unique_rows"], 1) + self.assertEqual(exposure["natural"]["label_draws"]["b"], 0) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/tests/test_token_budget.py b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_token_budget.py new file mode 100644 index 0000000000000000000000000000000000000000..1c685c0a64f5d53e85de9f9a833c1193dc7cd216 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/tests/test_token_budget.py @@ -0,0 +1,132 @@ +import copy +import importlib.util +import random +import sys +import unittest +from collections import Counter +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parents[5])) +from src.training.model_classifier.sequence_repair.train import ( + global_batch_loss, + sample_source_balanced, + source_sampling_pools, + token_budget_microbatches, +) + + +class TokenBudgetTests(unittest.TestCase): + def test_stable_groups_preserve_duplicate_draws_and_bound_padding(self): + lengths = [3, 12, 32, 7, 3] + sampled = [2, 4, 0, 1, 2, 3, 4, 1] + groups = token_budget_microbatches(sampled, lengths, 32) + flat = [index for group in groups for index in group] + self.assertEqual(Counter(flat), Counter(sampled)) + self.assertEqual(flat, sorted(sampled, key=lengths.__getitem__)) + longest_index = 2 + self.assertEqual( + [group for group in groups if longest_index in group], + [[longest_index], [longest_index]], + ) + for group in groups: + self.assertLessEqual(len(group) * max(lengths[i] for i in group), 32) + with self.assertRaises(ValueError): + token_budget_microbatches(sampled, lengths, 31) + with self.assertRaises(ValueError): + token_budget_microbatches(sampled, lengths, 0) + + def test_prefetch_preserves_sampling_and_rng_across_optimizer_steps(self): + rows = [ + {"source": source, "length_bucket": length, "label": label} + for source in ["natural", "authored"] + for length in [8, 32] + for label in ["a", "b"] + ] + pools = source_sampling_pools(rows, True) + sequential, prefetched = random.Random(17), random.Random(17) + for _ in range(30): + original = [] + for _ in range(2): + original.extend( + sample_source_balanced(pools, sequential, True) for _ in range(4) + ) + sampled = [ + sample_source_balanced(pools, prefetched, True) for _ in range(8) + ] + self.assertEqual(original, sampled) + groups = token_budget_microbatches( + sampled, [r["length_bucket"] for r in rows], 32 + ) + self.assertEqual( + Counter(original), Counter(i for group in groups for i in group) + ) + self.assertEqual(sequential.getstate(), prefetched.getstate()) + + @unittest.skipUnless( + importlib.util.find_spec("torch") and importlib.util.find_spec("transformers"), + "Optional actual ModernBERT gradient check requires Torch and Transformers", + ) + def test_fp32_modernbert_gradient_matches_full_batch_with_dropout_disabled(self): + import torch # noqa: PLC0415 + from transformers import ( # noqa: PLC0415 - optional dependency + ModernBertConfig, + ModernBertForSequenceClassification, + ) + + torch.set_num_threads(1) + torch.manual_seed(11) + config = ModernBertConfig( + vocab_size=64, + hidden_size=16, + intermediate_size=32, + num_hidden_layers=2, + num_attention_heads=2, + max_position_embeddings=64, + local_attention=8, + global_attn_every_n_layers=2, + num_labels=3, + pad_token_id=0, + cls_token_id=1, + sep_token_id=2, + embedding_dropout=0.0, + attention_dropout=0.0, + mlp_dropout=0.0, + classifier_dropout=0.0, + classifier_pooling="mean", + reference_compile=False, + ) + config._attn_implementation = "sdpa" + full = ModernBertForSequenceClassification(config).float().train() + grouped = copy.deepcopy(full) + lengths = [3, 12, 32, 7, 5, 19, 8, 4] + examples = [torch.randint(3, 64, (length,)) for length in lengths] + labels = torch.arange(8) % 3 + + def accumulate(model, indices): + ids = torch.zeros( + len(indices), max(lengths[i] for i in indices), dtype=torch.long + ) + mask = torch.zeros_like(ids) + for j, index in enumerate(indices): + ids[j, : lengths[index]] = examples[index] + mask[j, : lengths[index]] = 1 + logits = model(input_ids=ids, attention_mask=mask).logits + global_batch_loss(logits, labels[indices], 8).backward() + + accumulate(full, list(range(8))) + groups = token_budget_microbatches(list(range(8)), lengths, 32) + self.assertGreater(len(groups), 2) + for indices in groups: + accumulate(grouped, indices) + for (name, expected), (_, actual) in zip( + full.named_parameters(), grouped.named_parameters(), strict=True + ): + self.assertIsNotNone(expected.grad, name) + self.assertIsNotNone(actual.grad, name) + torch.testing.assert_close( + expected.grad, actual.grad, rtol=2e-5, atol=2e-6, msg=name + ) + + +if __name__ == "__main__": + unittest.main() diff --git a/reproducibility/src/training/model_classifier/sequence_repair/train.py b/reproducibility/src/training/model_classifier/sequence_repair/train.py new file mode 100644 index 0000000000000000000000000000000000000000..f703eeaecc61780daab1f3ff74dac47b0c108252 --- /dev/null +++ b/reproducibility/src/training/model_classifier/sequence_repair/train.py @@ -0,0 +1,526 @@ +"""Explicit single-GPU LoRA continuation with development-only selection.""" + +# ruff: noqa: PLC0415 + +import argparse +import hashlib +import json +import math +import random +import time +from collections import Counter, defaultdict +from pathlib import Path + +from .data import assert_disjoint, file_receipts, label_counts, read_records +from .evaluate import evaluate_records +from .model import load_model, save_adapter + + +def microbatch_loss(logits, labels, accumulation_steps): + from torch.nn import functional + + if accumulation_steps <= 0: + raise ValueError("Gradient accumulation must be positive") + # Each microbatch has the same number of examples. HF Trainer loss-kwargs + # normalization is intentionally not involved in this explicit loop. + return ( + functional.cross_entropy(logits.float(), labels, reduction="mean") + / accumulation_steps + ) + + +def global_batch_loss(logits, labels, examples_per_step): + """Normalize variable microbatches by the same sampled optimizer-step size.""" + from torch.nn import functional + + if examples_per_step <= 0: + raise ValueError("Optimizer-step example count must be positive") + return ( + functional.cross_entropy(logits.float(), labels, reduction="sum") + / examples_per_step + ) + + +def token_budget_microbatches(indices, lengths, token_budget): + """Stable length ordering without dropping or duplicating sampled examples.""" + if token_budget <= 0: + raise ValueError("Microbatch token budget must be positive") + batches, current = [], [] + for index in sorted(indices, key=lengths.__getitem__): + length = lengths[index] + if length > token_budget: + raise ValueError("A sampled example exceeds the microbatch token budget") + if current and length * (len(current) + 1) > token_budget: + batches.append(current) + current = [] + current.append(index) + if current: + batches.append(current) + return batches + + +def selection_score(metrics, selection): + if selection == "macro-f1": + return metrics["macro_f1"] + if selection in { + "length-macro-f1", + "source-macro-f1", + "source-present-macro-f1", + }: + field = "length_bucket" if selection == "length-macro-f1" else "source" + metric = ( + "macro_f1_present_labels" + if selection == "source-present-macro-f1" + else "macro_f1" + ) + groups = list(metrics["breakdowns"][field].values()) + if not groups: + raise ValueError("Grouped selection requires measured groups") + return sum(group[metric] for group in groups) / len(groups) + raise ValueError("Unknown development selection metric") + + +def length_sampling_pools(rows): + buckets = defaultdict(lambda: defaultdict(list)) + for index, row in enumerate(rows): + buckets[str(row.get("length_bucket", "unspecified"))][row["label"]].append( + index + ) + return [list(labels.values()) for labels in buckets.values()] + + +def sample_length_balanced(pools, rng, balance_labels): + bucket = rng.choice(pools) + pool = ( + rng.choice(bucket) + if balance_labels + else [index for label in bucket for index in label] + ) + return rng.choice(pool) + + +def source_sampling_pools(rows, balance_lengths): + sources = defaultdict(lambda: defaultdict(lambda: defaultdict(list))) + for index, row in enumerate(rows): + if not row.get("source"): + raise ValueError("Source-balanced sampling requires explicit sources") + length = ( + str(row.get("length_bucket", "unspecified")) if balance_lengths else "all" + ) + sources[row["source"]][length][row["label"]].append(index) + return [ + [list(labels.values()) for labels in lengths.values()] + for lengths in sources.values() + ] + + +def source_sampling_weights(rows, specification): + """Resolve explicit probabilities in the same order as source pools.""" + if specification is None: + return None + weights = json.loads(specification) + sources = list(dict.fromkeys(row.get("source") for row in rows)) + if ( + not sources + or not all(sources) + or not isinstance(weights, dict) + or set(weights) != set(sources) + ): + raise ValueError("Source weights must name every eligible source exactly") + values = [weights[source] for source in sources] + if any( + isinstance(value, bool) + or not isinstance(value, (int, float)) + or not math.isfinite(value) + or value <= 0 + for value in values + ): + raise ValueError("Source weights must be finite positive numbers") + total = sum(values) + if not math.isfinite(total): + raise ValueError("Total source weight must be finite") + return [value / total for value in values] + + +def sample_source_balanced(pools, rng, balance_labels, source_weights=None): + source = ( + rng.choice(pools) + if source_weights is None + else rng.choices(pools, weights=source_weights, k=1)[0] + ) + return sample_length_balanced(source, rng, balance_labels) + + +def sampling_exposure(rows, drawn): + """Report sampling counts separately from class/length selection metrics.""" + sources = defaultdict( + lambda: {"rows": 0, "draws": 0, "unique_rows": 0, "label_draws": Counter()} + ) + for index, row in enumerate(rows): + item = sources[row.get("source", "unspecified")] + count = drawn[index] + item["rows"] += 1 + item["draws"] += count + item["unique_rows"] += int(count > 0) + item["label_draws"][row["label"]] += count + return dict(sources) + + +def main(): + import torch + + parser = argparse.ArgumentParser() + parser.add_argument("--base", required=True) + parser.add_argument("--base-id", default="llm-semantic-router/mmbert-32k-yarn") + parser.add_argument("--base-revision", required=True) + parser.add_argument("--adapter", type=Path, required=True) + parser.add_argument("--contract", required=True) + parser.add_argument("--train", nargs="+", required=True) + parser.add_argument("--dev", nargs="+", required=True) + parser.add_argument("--output", type=Path, required=True) + parser.add_argument("--steps", type=int, default=400) + parser.add_argument("--batch-size", type=int, default=8) + parser.add_argument("--accumulate", type=int, default=2) + parser.add_argument("--max-length", type=int, default=2048) + parser.add_argument("--learning-rate", type=float, default=2e-5) + parser.add_argument("--eval-every", type=int, default=100) + parser.add_argument( + "--selection", + choices=[ + "macro-f1", + "length-macro-f1", + "source-macro-f1", + "source-present-macro-f1", + ], + default="macro-f1", + ) + parser.add_argument("--balanced-sampling", action="store_true") + parser.add_argument("--length-balanced-sampling", action="store_true") + parser.add_argument("--source-balanced-sampling", action="store_true") + parser.add_argument( + "--source-weights", help="JSON object of explicit source weights" + ) + parser.add_argument("--microbatch-token-budget", type=int) + parser.add_argument("--probe-only", action="store_true") + parser.add_argument("--seed", type=int, default=20260913) + args = parser.parse_args() + if ( + min( + args.steps, + args.batch_size, + args.accumulate, + args.max_length, + args.eval_every, + ) + <= 0 + or args.learning_rate <= 0 + or ( + args.microbatch_token_budget is not None + and args.microbatch_token_budget <= 0 + ) + ): + raise ValueError("Training budgets and learning rate must be positive") + if args.output.exists() and any(args.output.iterdir()): + raise ValueError("Refusing to overwrite a training run") + torch.set_num_threads(8) + torch.manual_seed(args.seed) + rng = random.Random(args.seed) + model, tokenizer, label_to_id, id_to_label = load_model( + args.base, args.contract, args.adapter, trainable=True + ) + if model.config.problem_type != "single_label_classification": + raise ValueError("This training loop requires single-label targets") + if args.max_length > model.config.max_position_embeddings: + raise ValueError("Training budget exceeds checkpoint position capacity") + rows, dev = read_records(args.train, label_to_id), read_records( + args.dev, label_to_id + ) + assert_disjoint(rows, dev) + train, rejected, pools = [], [], defaultdict(list) + for row in rows: + encoded = tokenizer(row["text"], truncation=False, padding=False) + if len(encoded["input_ids"]) > args.max_length: + rejected.append( + { + "id": row["id"], + "reason": "explicit context budget", + "tokens": len(encoded["input_ids"]), + } + ) + continue + index = len(train) + train.append((row, encoded)) + pools[row["label"]].append(index) + if not train: + raise ValueError("No training rows fit the explicit budget") + train_lengths = [len(encoded["input_ids"]) for _, encoded in train] + if ( + args.microbatch_token_budget is not None + and max(train_lengths) > args.microbatch_token_budget + ): + raise ValueError( + "Microbatch token budget cannot fit the longest eligible example" + ) + pool_values = list(pools.values()) + length_pools = length_sampling_pools([row for row, _encoded in train]) + balance_sources = args.source_balanced_sampling or args.source_weights is not None + source_weights = source_sampling_weights( + [row for row, _encoded in train], args.source_weights + ) + source_pools = ( + source_sampling_pools( + [row for row, _encoded in train], args.length_balanced_sampling + ) + if balance_sources + else None + ) + model.gradient_checkpointing_enable( + gradient_checkpointing_kwargs={"use_reentrant": False} + ) + model.to("cuda") + model.train() + parameters = [ + parameter for parameter in model.parameters() if parameter.requires_grad + ] + optimizer = torch.optim.AdamW(parameters, lr=args.learning_rate, weight_decay=0.01) + args.output.mkdir(parents=True, exist_ok=True) + metadata = { + "base_model": args.base_id, + "base_revision": args.base_revision, + "initial_adapter_sha256": hashlib.sha256( + (args.adapter / "adapter_model.safetensors").read_bytes() + ).hexdigest(), + "train_files": file_receipts(args.train), + "dev_files": file_receipts(args.dev), + "label2id": label_to_id, + "train_rows": len(train), + "dev_rows": len(dev), + "train_label_counts": label_counts([row for row, _ in train]), + "rejected": rejected, + "seed": args.seed, + "steps": args.steps, + "batch_size": args.batch_size, + "accumulate": args.accumulate, + "examples_per_optimizer_step": args.batch_size * args.accumulate, + "microbatch_token_budget": args.microbatch_token_budget, + "dropout_trajectory_equivalence": args.microbatch_token_budget is None, + "max_length": args.max_length, + "learning_rate": args.learning_rate, + "balanced_sampling": args.balanced_sampling, + "length_balanced_sampling": args.length_balanced_sampling, + "source_balanced_sampling": balance_sources, + "source_sampling_probabilities": ( + dict( + zip( + dict.fromkeys(row["source"] for row, _ in train), + source_weights, + strict=True, + ) + ) + if source_weights is not None + else None + ), + "implementation_sha256": hashlib.sha256( + Path(__file__).read_bytes() + ).hexdigest(), + "selection": args.selection, + "test_used": False, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": sum(parameter.numel() for parameter in model.parameters()), + "trainable_parameters": sum(parameter.numel() for parameter in parameters), + } + (args.output / "run.json").write_text(json.dumps(metadata, indent=2) + "\n") + print( + json.dumps( + { + "event": "ready", + "train_rows": len(train), + "rejected_rows": len(rejected), + "trainable_parameters": metadata["trainable_parameters"], + } + ), + flush=True, + ) + best = -1.0 + drawn = Counter() + + def evaluate_checkpoint(step): + nonlocal best + metrics, _predictions = evaluate_records( + model, tokenizer, dev, label_to_id, id_to_label, args.max_length + ) + score = selection_score(metrics, args.selection) + (args.output / f"sampling-step-{step}.json").write_text( + json.dumps(sampling_exposure([row for row, _ in train], drawn), indent=2) + + "\n" + ) + (args.output / f"dev-step-{step}.json").write_text( + json.dumps(metrics, indent=2) + "\n" + ) + print( + json.dumps( + { + "event": "dev", + "step": step, + "score": score, + "accuracy": metrics["accuracy"], + "macro_f1": metrics["macro_f1"], + } + ), + flush=True, + ) + if score > best: + best = score + save_adapter( + model, + tokenizer, + args.output / "best-adapter", + args.base_id, + args.base_revision, + ) + (args.output / "selection.json").write_text( + json.dumps( + { + "step": step, + "score": score, + "metric": args.selection, + "test_used": False, + }, + indent=2, + ) + + "\n" + ) + + def draw_microbatch_indices(): + return [ + ( + sample_source_balanced( + source_pools, rng, args.balanced_sampling, source_weights + ) + if balance_sources + else ( + sample_length_balanced(length_pools, rng, args.balanced_sampling) + if args.length_balanced_sampling + else ( + rng.choice(rng.choice(pool_values)) + if args.balanced_sampling + else rng.randrange(len(train)) + ) + ) + ) + for _ in range(args.batch_size) + ] + + if not args.probe_only: + evaluate_checkpoint(0) + for step in range(1, args.steps + 1): + optimizer.zero_grad(set_to_none=True) + torch.cuda.reset_peak_memory_stats() + torch.cuda.synchronize() + start, step_loss, lengths = time.perf_counter(), 0.0, [] + if args.microbatch_token_budget is None: + microbatches = (draw_microbatch_indices() for _ in range(args.accumulate)) + else: + sampled = [ + index + for _ in range(args.accumulate) + for index in draw_microbatch_indices() + ] + microbatches = token_budget_microbatches( + sampled, train_lengths, args.microbatch_token_budget + ) + microbatch_count, max_padded_tokens = 0, 0 + for indices in microbatches: + microbatch_count += 1 + max_padded_tokens = max( + max_padded_tokens, max(train_lengths[i] for i in indices) * len(indices) + ) + drawn.update(indices) + selected = [train[index] for index in indices] + lengths.extend(len(encoded["input_ids"]) for _, encoded in selected) + batch = tokenizer.pad( + [encoded for _, encoded in selected], padding=True, return_tensors="pt" + ) + batch = {key: value.to("cuda") for key, value in batch.items()} + gold = torch.tensor( + [label_to_id[row["label"]] for row, _ in selected], device="cuda" + ) + with torch.autocast("cuda", dtype=torch.bfloat16): + logits = model(**batch).logits + loss = ( + microbatch_loss(logits, gold, args.accumulate) + if args.microbatch_token_budget is None + else global_batch_loss( + logits, gold, args.batch_size * args.accumulate + ) + ) + if not bool(torch.isfinite(loss)): + raise ValueError(f"Non-finite loss at step {step}") + loss.backward() + step_loss += float(loss.detach()) + del logits, loss, batch + gradients = [ + parameter.grad for parameter in parameters if parameter.grad is not None + ] + if not gradients or not all( + bool(torch.isfinite(gradient).all()) for gradient in gradients + ): + raise ValueError(f"Missing or non-finite gradients at step {step}") + norm = torch.nn.utils.clip_grad_norm_(parameters, 1.0) + warmup = max(1, round(args.steps * 0.06)) + rate = ( + step / warmup + if step <= warmup + else 0.5 + * (1 + math.cos(math.pi * (step - warmup) / max(1, args.steps - warmup))) + ) + for group in optimizer.param_groups: + group["lr"] = args.learning_rate * rate + optimizer.step() + torch.cuda.synchronize() + log = { + "step": step, + "loss": step_loss, + "grad_norm": float(norm), + "finite": True, + "gradient_tensors": len(gradients), + "max_tokens": max(lengths), + "mean_tokens": sum(lengths) / len(lengths), + "microbatches": microbatch_count, + "max_padded_tokens": max_padded_tokens, + "step_seconds": time.perf_counter() - start, + "peak_allocated_gib": torch.cuda.max_memory_allocated() / 2**30, + "peak_reserved_gib": torch.cuda.max_memory_reserved() / 2**30, + } + with (args.output / "steps.jsonl").open("a") as stream: + stream.write(json.dumps(log) + "\n") + if step == 1 or step % 10 == 0 or args.probe_only: + print(json.dumps(log), flush=True) + if not args.probe_only and (step % args.eval_every == 0 or step == args.steps): + evaluate_checkpoint(step) + (args.output / "sampling-final.json").write_text( + json.dumps(sampling_exposure([row for row, _ in train], drawn), indent=2) + "\n" + ) + if not args.probe_only: + save_adapter( + model, + tokenizer, + args.output / "last-adapter", + args.base_id, + args.base_revision, + ) + print( + json.dumps( + { + "event": "complete", + "best_dev_score": best if not args.probe_only else None, + } + ), + flush=True, + ) + + +if __name__ == "__main__": + main() diff --git a/reproducibility/summarize.py b/reproducibility/summarize.py new file mode 100644 index 0000000000000000000000000000000000000000..b335247b7956fda9fdf54bc48d65ee674e5baa5b --- /dev/null +++ b/reproducibility/summarize.py @@ -0,0 +1,27 @@ +"""Recompute Modality source/language/length comparisons from frozen predictions.""" +import argparse,hashlib,json +from pathlib import Path +from src.training.model_classifier.sequence_repair.data import classification_metrics +ALIASES={'eng':'en','zho':'zh','spa':'es','fra':'fr','deu':'de','jpn':'ja'} + +def records(path):return [json.loads(x) for x in path.read_text().split('\n') if x] +def main(): + p=argparse.ArgumentParser();p.add_argument('--manifest',type=Path,required=True);p.add_argument('--baseline',type=Path,required=True);p.add_argument('--candidate',type=Path,required=True);p.add_argument('--contract',type=Path,required=True);p.add_argument('--output',type=Path,required=True);a=p.parse_args() + if a.output.exists():raise ValueError('Refusing to replace an existing report') + rows=records(a.manifest);preds={name:records(getattr(a,name)) for name in ['baseline','candidate']};contract=json.loads(a.contract.read_text());labels=contract['label2id'];ids={int(k):v for k,v in contract['id2label'].items()} + for name,values in preds.items(): + if len(values)!=len(rows) or len({x['id'] for x in values})!=len(rows):raise ValueError('Prediction rows differ') + if [x['id'] for x in values]!=[x['id'] for x in rows]:raise ValueError('Prediction order differs') + def compare(indices): + gold=[labels[rows[i]['label']] for i in indices];result={'rows':len(indices),'source_groups':len({rows[i]['group_id'] for i in indices})} + for name,values in preds.items():result[name]=classification_metrics(gold,[values[i]['prediction'] for i in indices],ids) + result['paired_correct_changes']={'improved':sum(preds['baseline'][i]['prediction']!=labels[rows[i]['label']] and preds['candidate'][i]['prediction']==labels[rows[i]['label']] for i in indices),'regressed':sum(preds['baseline'][i]['prediction']==labels[rows[i]['label']] and preds['candidate'][i]['prediction']!=labels[rows[i]['label']] for i in indices)} + return result + report={'selection_used_test':False,'metric':'Argmax classification on FP32 frozen predictions; academic and authored stress are not natural-request estimates','inputs':{key:hashlib.sha256(getattr(a,key).read_bytes()).hexdigest() for key in ['manifest','baseline','candidate','contract']},'sources':{}} + for source in sorted({x['source'] for x in rows}): + indices=[i for i,x in enumerate(rows) if x['source']==source];result=compare(indices) + result['by_language']={language:compare([i for i in indices if ALIASES.get(rows[i]['language'],rows[i]['language'])==language]) for language in sorted({ALIASES.get(rows[i]['language'],rows[i]['language']) for i in indices})} + result['by_length']={length:compare([i for i in indices if rows[i]['length_bucket']==length]) for length in sorted({rows[i]['length_bucket'] for i in indices})} + report['sources'][source]=result + a.output.write_text(json.dumps(report,indent=2)+'\n') +if __name__=='__main__':main() diff --git a/reproducibility/train_stages.py b/reproducibility/train_stages.py new file mode 100644 index 0000000000000000000000000000000000000000..974e661c0fdeb19cf6c55be465fc030b37f1552b --- /dev/null +++ b/reproducibility/train_stages.py @@ -0,0 +1,32 @@ +"""Selected Modality curriculum with separate frozen export and final evaluation.""" +import argparse,hashlib,json,subprocess,sys +from pathlib import Path +BASE_ID='llm-semantic-router/mmbert-32k-yarn';BASE_REV='72a23a6640489471eb4ff7ad3ec5bc80af8a27de' +VELA_ID='llm-semantic-router/Vela-1.0-Encoder-307M';VELA_REV='5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114' +STAGES=['modality-v1','modality-v2-diversity','modality-v4-diverse-4k','modality-v5-diverse-context','modality-v6-vela-base'] + +def main(): + p=argparse.ArgumentParser();p.add_argument('--stage',choices=['initialize',*STAGES,'export'],required=True);p.add_argument('--artifacts',type=Path,required=True);a=p.parse_args();r=a.artifacts;module='src.training.model_classifier.sequence_repair.' + contract=r/'data/modality-v1/contract.json' + if a.stage=='initialize': + cmd=[sys.executable,'-m',module+'initialize','--base',str(r/'base'),'--base-id',BASE_ID,'--base-revision',BASE_REV,'--contract',str(contract),'--seed','20260913','--output',str(r/'initial-adapter')] + elif a.stage=='export': + lineage=json.loads(Path('release-lineage.json').read_text());stage=lineage['final_stage'];vela=stage=='modality-v6-vela-base' + cmd=[sys.executable,'-m',module+'export','--base',str(r/('vela-base' if vela else 'base')),'--base-id',VELA_ID if vela else BASE_ID,'--base-revision',VELA_REV if vela else BASE_REV,'--adapter',str(r/'runs'/stage/'best-adapter'),'--contract',str(contract),'--output',str(r/'frozen'),'--run-manifest',str(r/'runs'/stage/'run.json'),'--runtime-task','modality'] + else: + i=STAGES.index(a.stage);receipt=json.loads((Path('training-receipts')/a.stage/'run.json').read_text());vela=i==4 + short='modality-v1' if i==0 else 'modality-v2';train=[r/'data'/short/'train.jsonl'];dev=[r/'data'/short/'dev.jsonl'] + if i>=2: + context='modality-v2-context-4k-v2' if i==2 else 'modality-v2-context-v2';train.append(r/'data'/context/'train.jsonl');dev.append(r/'data'/context/'dev.jsonl') + for role,files in [('train',train),('dev',dev)]: + expected=receipt[role+'_files'] + if len(expected)!=len(files):raise ValueError('Training file count differs') + for path,item in zip(files,expected): + if hashlib.sha256(path.read_bytes()).hexdigest()!=item['sha256']:raise ValueError('Data checksum differs: '+str(path)) + adapter=r/'initial-adapter' if i==0 else r/'runs'/STAGES[i-1]/'best-adapter' + cmd=[sys.executable,'-m',module+'train','--base',str(r/('vela-base' if vela else 'base')),'--base-id',VELA_ID if vela else BASE_ID,'--base-revision',VELA_REV if vela else BASE_REV,'--adapter',str(adapter),'--contract',str(contract),'--train',*map(str,train),'--dev',*map(str,dev),'--output',str(r/'runs'/a.stage),'--steps',str(receipt['steps']),'--batch-size',str(receipt['batch_size']),'--accumulate',str(receipt['accumulate']),'--max-length',str(receipt['max_length']),'--learning-rate',str(receipt['learning_rate']),'--seed',str(receipt['seed']),'--eval-every','50' if vela else '100','--selection',receipt['selection']] + for key in ['balanced_sampling','length_balanced_sampling','source_balanced_sampling']: + if receipt.get(key):cmd.append('--'+key.replace('_','-')) + if vela:cmd.extend(['--microbatch-token-budget','32768']) + print(json.dumps({'command':cmd,'test_used_for_selection':False}),flush=True);subprocess.run(cmd,check=True) +if __name__=='__main__':main() diff --git a/reproducibility/training-receipts/modality-v1/run.json b/reproducibility/training-receipts/modality-v1/run.json new file mode 100644 index 0000000000000000000000000000000000000000..adcffa20c3f8325810fd5703fcfe6202a52e61dc --- /dev/null +++ b/reproducibility/training-receipts/modality-v1/run.json @@ -0,0 +1,46 @@ +{ + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "initial_adapter_sha256": "c11b45350cdac82aafa2943397f73ccbc6e31c9ee7806cb4038cf50333b3434d", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "b92180b7cfb98d6232805f64ffd6448d74e6a82c23dd2c4310a0f803541a5ab8" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 4485, + "dev_rows": 929, + "train_label_counts": { + "AR": 2085, + "DIFFUSION": 1200, + "BOTH": 1200 + }, + "rejected": [], + "seed": 20260913, + "steps": 600, + "batch_size": 8, + "accumulate": 2, + "max_length": 2048, + "learning_rate": 3e-05, + "balanced_sampling": true, + "length_balanced_sampling": false, + "implementation_sha256": "8ac6622616f3054ff325de43ecdb9c6654762bb1813766a5cba8cdf274329304", + "selection": "macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v1/selection.json b/reproducibility/training-receipts/modality-v1/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..25c8059f61dc4960e8fb78520e6323b9dbf3d089 --- /dev/null +++ b/reproducibility/training-receipts/modality-v1/selection.json @@ -0,0 +1,6 @@ +{ + "step": 100, + "score": 0.648545274764787, + "metric": "macro-f1", + "test_used": false +} diff --git a/reproducibility/training-receipts/modality-v2-diversity/run.json b/reproducibility/training-receipts/modality-v2-diversity/run.json new file mode 100644 index 0000000000000000000000000000000000000000..e800eb0d0792bff2c955a713557d40bbcf9a054e --- /dev/null +++ b/reproducibility/training-receipts/modality-v2-diversity/run.json @@ -0,0 +1,47 @@ +{ + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "initial_adapter_sha256": "846f145e8c88560dfdbd87127f3fe4f88cf1109ea9b013c5540492ad8633afde", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 18885, + "dev_rows": 929, + "train_label_counts": { + "AR": 6885, + "DIFFUSION": 6000, + "BOTH": 6000 + }, + "rejected": [], + "seed": 20260913, + "steps": 600, + "batch_size": 8, + "accumulate": 2, + "max_length": 2048, + "learning_rate": 1e-05, + "balanced_sampling": true, + "length_balanced_sampling": false, + "source_balanced_sampling": true, + "implementation_sha256": "0de11b6ccec5cf1fc47f23c2cedfdcce2693f8905ebe778342aace8633f4416a", + "selection": "macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v2-diversity/selection.json b/reproducibility/training-receipts/modality-v2-diversity/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..3762a861984a6655580751f99b2509079ff64180 --- /dev/null +++ b/reproducibility/training-receipts/modality-v2-diversity/selection.json @@ -0,0 +1,6 @@ +{ + "step": 300, + "score": 0.8362344436378876, + "metric": "macro-f1", + "test_used": false +} diff --git a/reproducibility/training-receipts/modality-v3-context/run.json b/reproducibility/training-receipts/modality-v3-context/run.json new file mode 100644 index 0000000000000000000000000000000000000000..d5db5a1155b44647d505c8a2eec8d6cbe566e5cd --- /dev/null +++ b/reproducibility/training-receipts/modality-v3-context/run.json @@ -0,0 +1,55 @@ +{ + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "initial_adapter_sha256": "4b423278960dc1048bdbbea44b39120480e74008cadea676286f68fec3da7719", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + }, + { + "file": "train.jsonl", + "sha256": "3ff3dadbbd665e072ae514b1d8590af4001504608b68e6992f8ccae7414007c4" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 19101, + "dev_rows": 1001, + "train_label_counts": { + "AR": 6957, + "DIFFUSION": 6072, + "BOTH": 6072 + }, + "rejected": [], + "seed": 20260913, + "steps": 400, + "batch_size": 1, + "accumulate": 4, + "max_length": 32768, + "learning_rate": 8e-06, + "balanced_sampling": true, + "length_balanced_sampling": true, + "source_balanced_sampling": true, + "implementation_sha256": "0de11b6ccec5cf1fc47f23c2cedfdcce2693f8905ebe778342aace8633f4416a", + "selection": "source-macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v3-context/selection.json b/reproducibility/training-receipts/modality-v3-context/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..83308e33350c6c9746d9fef4f30b251684f3fe3f --- /dev/null +++ b/reproducibility/training-receipts/modality-v3-context/selection.json @@ -0,0 +1,6 @@ +{ + "step": 400, + "score": 0.4892126029085696, + "metric": "source-macro-f1", + "test_used": false +} diff --git a/reproducibility/training-receipts/modality-v4-diverse-4k/run.json b/reproducibility/training-receipts/modality-v4-diverse-4k/run.json new file mode 100644 index 0000000000000000000000000000000000000000..4a67e2d5289f75413a673a26d0d49fb0f82ebcca --- /dev/null +++ b/reproducibility/training-receipts/modality-v4-diverse-4k/run.json @@ -0,0 +1,55 @@ +{ + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "initial_adapter_sha256": "4b423278960dc1048bdbbea44b39120480e74008cadea676286f68fec3da7719", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + }, + { + "file": "train.jsonl", + "sha256": "9c02fd4abfc08e29a3fc1f207157631f65f2f24ecaae5ea01b5d5f8b056f6ba3" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "f99b3cb0e2608f385944d0bcd4023bc611f49e7bab02857e29436d5d71a9f263" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 19101, + "dev_rows": 947, + "train_label_counts": { + "AR": 6957, + "DIFFUSION": 6072, + "BOTH": 6072 + }, + "rejected": [], + "seed": 20260913, + "steps": 600, + "batch_size": 1, + "accumulate": 4, + "max_length": 4096, + "learning_rate": 1e-05, + "balanced_sampling": true, + "length_balanced_sampling": true, + "source_balanced_sampling": false, + "implementation_sha256": "0de11b6ccec5cf1fc47f23c2cedfdcce2693f8905ebe778342aace8633f4416a", + "selection": "source-macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v4-diverse-4k/selection.json b/reproducibility/training-receipts/modality-v4-diverse-4k/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..8e794c7aaff51e9e7bf3ded65e46a73402e1f114 --- /dev/null +++ b/reproducibility/training-receipts/modality-v4-diverse-4k/selection.json @@ -0,0 +1,6 @@ +{ + "step": 400, + "score": 0.5671738538211304, + "metric": "source-macro-f1", + "test_used": false +} diff --git a/reproducibility/training-receipts/modality-v5-diverse-context/run.json b/reproducibility/training-receipts/modality-v5-diverse-context/run.json new file mode 100644 index 0000000000000000000000000000000000000000..098c016a695b4cabaa20803733954469fe121531 --- /dev/null +++ b/reproducibility/training-receipts/modality-v5-diverse-context/run.json @@ -0,0 +1,55 @@ +{ + "base_model": "llm-semantic-router/mmbert-32k-yarn", + "base_revision": "72a23a6640489471eb4ff7ad3ec5bc80af8a27de", + "initial_adapter_sha256": "cdc09cfea783757d1f4a2a4dcdfa48dfe75bf3dd228f4049da6340e9e357946e", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + }, + { + "file": "train.jsonl", + "sha256": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 19749, + "dev_rows": 1001, + "train_label_counts": { + "AR": 7173, + "DIFFUSION": 6288, + "BOTH": 6288 + }, + "rejected": [], + "seed": 20260913, + "steps": 400, + "batch_size": 1, + "accumulate": 4, + "max_length": 32768, + "learning_rate": 5e-06, + "balanced_sampling": true, + "length_balanced_sampling": true, + "source_balanced_sampling": false, + "implementation_sha256": "0de11b6ccec5cf1fc47f23c2cedfdcce2693f8905ebe778342aace8633f4416a", + "selection": "source-macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v5-diverse-context/selection.json b/reproducibility/training-receipts/modality-v5-diverse-context/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..c89d60f2d9d0fe4ea5102e63d5fe7e01621588ae --- /dev/null +++ b/reproducibility/training-receipts/modality-v5-diverse-context/selection.json @@ -0,0 +1,6 @@ +{ + "step": 400, + "score": 0.5700111329868457, + "metric": "source-macro-f1", + "test_used": false +} diff --git a/reproducibility/training-receipts/modality-v6-vela-base/dev-step-0.json b/reproducibility/training-receipts/modality-v6-vela-base/dev-step-0.json new file mode 100644 index 0000000000000000000000000000000000000000..d0de95cc4f76b9c700fa7d6dbe961e14aa273866 --- /dev/null +++ b/reproducibility/training-receipts/modality-v6-vela-base/dev-step-0.json @@ -0,0 +1,1400 @@ +{ + "rows": 1001, + "accuracy": 0.8411588411588412, + "macro_f1": 0.8456991326446053, + "macro_f1_present_labels": 0.8456991326446053, + "per_label": { + "AR": { + "support": 473, + "tp": 314, + "fp": 0, + "fn": 159, + "precision": 1.0, + "recall": 0.6638477801268499, + "f1": 0.7979669631512071 + }, + "DIFFUSION": { + "support": 264, + "tp": 264, + "fp": 66, + "fn": 0, + "precision": 0.8, + "recall": 1.0, + "f1": 0.8888888888888888 + }, + "BOTH": { + "support": 264, + "tp": 264, + "fp": 93, + "fn": 0, + "precision": 0.7394957983193278, + "recall": 1.0, + "f1": 0.8502415458937198 + } + }, + "confusion": [ + [ + 314, + 66, + 93 + ], + [ + 0, + 264, + 0 + ], + [ + 0, + 0, + 264 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 180, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 180, + "tp": 180, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 180, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 24, + "tp": 24, + "fp": 12, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 0, + 24, + 0 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 720, + "accuracy": 0.7958333333333333, + "macro_f1": 0.7650437660438248, + "macro_f1_present_labels": 0.7650437660438248, + "per_label": { + "AR": { + "support": 240, + "tp": 93, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.3875, + "f1": 0.5585585585585585 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 93, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + }, + "aya-original-human": { + "rows": 29, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 29, + "tp": 29, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 29, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 150, + "accuracy": 0.9666666666666667, + "macro_f1": 0.9680464778503994, + "macro_f1_present_labels": 0.9680464778503994, + "per_label": { + "AR": { + "support": 70, + "tp": 65, + "fp": 0, + "fn": 5, + "precision": 1.0, + "recall": 0.9285714285714286, + "f1": 0.9629629629629629 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 5, + "fn": 0, + "precision": 0.8888888888888888, + "recall": 1.0, + "f1": 0.9411764705882353 + } + }, + "confusion": [ + [ + 65, + 0, + 5 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "deu": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 186, + "accuracy": 0.7903225806451613, + "macro_f1": 0.8009399198931909, + "macro_f1_present_labels": 0.8009399198931909, + "per_label": { + "AR": { + "support": 82, + "tp": 43, + "fp": 0, + "fn": 39, + "precision": 1.0, + "recall": 0.524390243902439, + "f1": 0.688 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 36, + "fn": 0, + "precision": 0.5909090909090909, + "recall": 1.0, + "f1": 0.7428571428571429 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 3, + "fn": 0, + "precision": 0.9454545454545454, + "recall": 1.0, + "f1": 0.9719626168224299 + } + }, + "confusion": [ + [ + 43, + 36, + 3 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "eng": { + "rows": 5, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 5, + "tp": 5, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 5, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 150, + "accuracy": 0.7333333333333333, + "macro_f1": 0.751191952597492, + "macro_f1_present_labels": 0.751191952597492, + "per_label": { + "AR": { + "support": 70, + "tp": 30, + "fp": 0, + "fn": 40, + "precision": 1.0, + "recall": 0.42857142857142855, + "f1": 0.6 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 2, + "fn": 0, + "precision": 0.9523809523809523, + "recall": 1.0, + "f1": 0.975609756097561 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 38, + "fn": 0, + "precision": 0.5128205128205128, + "recall": 1.0, + "f1": 0.6779661016949152 + } + }, + "confusion": [ + [ + 30, + 2, + 38 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "fr": { + "rows": 150, + "accuracy": 0.76, + "macro_f1": 0.7811671087533156, + "macro_f1_present_labels": 0.7811671087533156, + "per_label": { + "AR": { + "support": 70, + "tp": 34, + "fp": 0, + "fn": 36, + "precision": 1.0, + "recall": 0.4857142857142857, + "f1": 0.6538461538461539 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 36, + "fn": 0, + "precision": 0.5263157894736842, + "recall": 1.0, + "f1": 0.6896551724137931 + } + }, + "confusion": [ + [ + 34, + 0, + 36 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "ja": { + "rows": 150, + "accuracy": 0.9333333333333333, + "macro_f1": 0.9373219373219372, + "macro_f1_present_labels": 0.9373219373219372, + "per_label": { + "AR": { + "support": 70, + "tp": 60, + "fp": 0, + "fn": 10, + "precision": 1.0, + "recall": 0.8571428571428571, + "f1": 0.9230769230769231 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 10, + "fn": 0, + "precision": 0.8, + "recall": 1.0, + "f1": 0.8888888888888888 + } + }, + "confusion": [ + [ + 60, + 0, + 10 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "spa": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 186, + "accuracy": 0.8440860215053764, + "macro_f1": 0.8545133878467212, + "macro_f1_present_labels": 0.8545133878467212, + "per_label": { + "AR": { + "support": 82, + "tp": 53, + "fp": 0, + "fn": 29, + "precision": 1.0, + "recall": 0.6463414634146342, + "f1": 0.7851851851851852 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 28, + "fn": 0, + "precision": 0.65, + "recall": 1.0, + "f1": 0.7878787878787878 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 1, + "fn": 0, + "precision": 0.9811320754716981, + "recall": 1.0, + "f1": 0.9904761904761905 + } + }, + "confusion": [ + [ + 53, + 28, + 1 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "zho": { + "rows": 3, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 3, + "tp": 3, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "short": { + "rows": 929, + "accuracy": 0.8417653390742734, + "macro_f1": 0.8469445749752597, + "macro_f1_present_labels": 0.8469445749752597, + "per_label": { + "AR": { + "support": 449, + "tp": 302, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.6726057906458798, + "f1": 0.8042609853528628 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 302, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "unspecified": { + "rows": 929, + "accuracy": 0.8417653390742734, + "macro_f1": 0.8469445749752597, + "macro_f1_present_labels": 0.8469445749752597, + "per_label": { + "AR": { + "support": 449, + "tp": 302, + "fp": 0, + "fn": 147, + "precision": 1.0, + "recall": 0.6726057906458798, + "f1": 0.8042609853528628 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 54, + "fn": 0, + "precision": 0.8163265306122449, + "recall": 1.0, + "f1": 0.898876404494382 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 93, + "fn": 0, + "precision": 0.7207207207207207, + "recall": 1.0, + "f1": 0.837696335078534 + } + }, + "confusion": [ + [ + 302, + 54, + 93 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + } + }, + "actual_tokens": { + "min": 9, + "max": 32768 + }, + "forward_batches_ms": [ + 7951.3892399845645, + 24.682334042154253, + 22.049822029657662, + 23.360353079624474, + 24.3825699435547, + 24.256629985757172, + 34.24256504513323, + 24.72389896865934, + 25.427454966120422, + 25.752547080628574, + 25.04231804050505, + 26.008810964412987, + 24.839375051669776, + 25.79350199084729, + 25.620796950533986, + 26.337529066950083, + 25.452626985497773, + 25.084492983296514, + 23.373947013169527, + 24.835766991600394, + 23.670071037486196, + 24.040832999162376, + 23.82285601925105, + 25.718325981870294, + 23.94235203973949, + 67.28544598445296, + 24.240268976427615, + 23.970600916072726, + 23.203744902275503, + 23.825343931093812, + 23.9079010207206, + 23.8074540393427, + 23.921990068629384, + 23.63455598242581, + 23.618023958988488, + 24.021278019063175, + 22.90464600082487, + 24.535832926630974, + 24.153056088835, + 23.442632984369993, + 23.437032010406256, + 24.295803043060005, + 27.833965956233442, + 23.607409093528986, + 22.930328967049718, + 42.14263102039695, + 24.301220080815256, + 23.736124974675477, + 23.23442103806883, + 25.575098930858076, + 24.317973060533404, + 23.851417005062103, + 40.955358068458736, + 23.855107952840626, + 24.231220013462007, + 24.493156000971794, + 26.808291906490922, + 27.43015589658171, + 26.958324015140533, + 71.889009908773, + 72.07445299718529, + 77.03507505357265, + 71.22085092123598, + 67.19580304343253, + 67.25443690083921, + 65.86683902423829, + 63.32439510151744, + 69.9001990724355, + 68.7988098943606, + 68.84922296740115, + 69.11942001897842, + 116.24330107588321, + 73.93188297282904, + 67.55383894778788, + 74.96135798282921, + 72.42749794386327, + 85.383080993779, + 264.967464026995, + 233.73422503937036, + 217.2980859177187, + 210.3603979339823, + 225.80570506397635, + 231.01699200924486, + 241.72410590108484, + 243.69510693941265, + 259.3330650124699, + 238.79560793284327, + 229.84719602391124, + 214.14722898043692, + 204.2199659626931, + 232.0609879679978, + 238.42367192264646, + 242.08824604284018, + 227.58120903745294, + 252.2601429373026, + 890.6667679548264, + 842.941373004578, + 908.7620580103248, + 893.4195600450039, + 926.120436983183, + 902.130710077472, + 924.324799911119, + 908.0960509600118, + 912.6980019500479, + 876.3914960436523, + 903.661935008131, + 953.8581649539992, + 929.6590070007369, + 952.2273340262473, + 852.6811179472134, + 914.2397129908204, + 881.5291019855067, + 893.0511620128527, + 3552.9280280461535, + 3378.174526966177, + 3313.6938179377466, + 3344.3720099749044, + 3512.887193937786, + 3454.985450953245, + 3743.8422010745853, + 3427.7674079639837, + 3369.901490979828, + 3522.408411023207, + 3421.936928993091, + 3568.5369269922376, + 3506.619893014431, + 3715.2604189468548, + 3415.8284560544416, + 3412.3947619227692, + 3407.8273070044816, + 3613.4438320295885 + ], + "precision": { + "parameters": "float32", + "forward_autocast": "bfloat16", + "device_type": "cuda" + } +} diff --git a/reproducibility/training-receipts/modality-v6-vela-base/dev-step-200.json b/reproducibility/training-receipts/modality-v6-vela-base/dev-step-200.json new file mode 100644 index 0000000000000000000000000000000000000000..ea43eb48f702ee809fa54c331c7c313f1077336d --- /dev/null +++ b/reproducibility/training-receipts/modality-v6-vela-base/dev-step-200.json @@ -0,0 +1,1400 @@ +{ + "rows": 1001, + "accuracy": 0.8601398601398601, + "macro_f1": 0.8640630057179847, + "macro_f1_present_labels": 0.8640630057179847, + "per_label": { + "AR": { + "support": 473, + "tp": 333, + "fp": 0, + "fn": 140, + "precision": 1.0, + "recall": 0.7040169133192389, + "f1": 0.826302729528536 + }, + "DIFFUSION": { + "support": 264, + "tp": 264, + "fp": 70, + "fn": 0, + "precision": 0.7904191616766467, + "recall": 1.0, + "f1": 0.882943143812709 + }, + "BOTH": { + "support": 264, + "tp": 264, + "fp": 70, + "fn": 0, + "precision": 0.7904191616766467, + "recall": 1.0, + "f1": 0.882943143812709 + } + }, + "confusion": [ + [ + 333, + 70, + 70 + ], + [ + 0, + 264, + 0 + ], + [ + 0, + 0, + 264 + ] + ], + "breakdowns": { + "source": { + "academic-text-answer": { + "rows": 180, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 180, + "tp": 180, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 180, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "authored-context-on-frozen-source": { + "rows": 72, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 24, + "tp": 12, + "fp": 0, + "fn": 12, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 24, + "tp": 24, + "fp": 12, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 24, + "tp": 24, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 12, + 12, + 0 + ], + [ + 0, + 24, + 0 + ], + [ + 0, + 0, + 24 + ] + ] + }, + "authored-output-contract-with-gallery-caption": { + "rows": 720, + "accuracy": 0.8222222222222222, + "macro_f1": 0.8004280725470316, + "macro_f1_present_labels": 0.8004280725470316, + "per_label": { + "AR": { + "support": 240, + "tp": 112, + "fp": 0, + "fn": 128, + "precision": 1.0, + "recall": 0.4666666666666667, + "f1": 0.6363636363636364 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 58, + "fn": 0, + "precision": 0.8053691275167785, + "recall": 1.0, + "f1": 0.8921933085501859 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 70, + "fn": 0, + "precision": 0.7741935483870968, + "recall": 1.0, + "f1": 0.8727272727272727 + } + }, + "confusion": [ + [ + 112, + 58, + 70 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + }, + "aya-original-human": { + "rows": 29, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 29, + "tp": 29, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 29, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "language": { + "de": { + "rows": 150, + "accuracy": 1.0, + "macro_f1": 1.0, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 70, + "tp": 70, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 70, + 0, + 0 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "deu": { + "rows": 2, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 2, + "tp": 2, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 2, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "en": { + "rows": 186, + "accuracy": 0.7956989247311828, + "macro_f1": 0.8088258471237194, + "macro_f1_present_labels": 0.8088258471237194, + "per_label": { + "AR": { + "support": 82, + "tp": 44, + "fp": 0, + "fn": 38, + "precision": 1.0, + "recall": 0.5365853658536586, + "f1": 0.6984126984126984 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 37, + "fn": 0, + "precision": 0.5842696629213483, + "recall": 1.0, + "f1": 0.7375886524822695 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 1, + "fn": 0, + "precision": 0.9811320754716981, + "recall": 1.0, + "f1": 0.9904761904761905 + } + }, + "confusion": [ + [ + 44, + 37, + 1 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "eng": { + "rows": 5, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 5, + "tp": 5, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 5, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "es": { + "rows": 150, + "accuracy": 0.7333333333333333, + "macro_f1": 0.7456095481670929, + "macro_f1_present_labels": 0.7456095481670929, + "per_label": { + "AR": { + "support": 70, + "tp": 30, + "fp": 0, + "fn": 40, + "precision": 1.0, + "recall": 0.42857142857142855, + "f1": 0.6 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 5, + "fn": 0, + "precision": 0.8888888888888888, + "recall": 1.0, + "f1": 0.9411764705882353 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 35, + "fn": 0, + "precision": 0.5333333333333333, + "recall": 1.0, + "f1": 0.6956521739130435 + } + }, + "confusion": [ + [ + 30, + 5, + 35 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "fr": { + "rows": 150, + "accuracy": 0.8066666666666666, + "macro_f1": 0.824227897622393, + "macro_f1_present_labels": 0.824227897622393, + "per_label": { + "AR": { + "support": 70, + "tp": 41, + "fp": 0, + "fn": 29, + "precision": 1.0, + "recall": 0.5857142857142857, + "f1": 0.7387387387387387 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 29, + "fn": 0, + "precision": 0.5797101449275363, + "recall": 1.0, + "f1": 0.7339449541284404 + } + }, + "confusion": [ + [ + 41, + 0, + 29 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "ja": { + "rows": 150, + "accuracy": 0.9666666666666667, + "macro_f1": 0.9680464778503994, + "macro_f1_present_labels": 0.9680464778503994, + "per_label": { + "AR": { + "support": 70, + "tp": 65, + "fp": 0, + "fn": 5, + "precision": 1.0, + "recall": 0.9285714285714286, + "f1": 0.9629629629629629 + }, + "DIFFUSION": { + "support": 40, + "tp": 40, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "BOTH": { + "support": 40, + "tp": 40, + "fp": 5, + "fn": 0, + "precision": 0.8888888888888888, + "recall": 1.0, + "f1": 0.9411764705882353 + } + }, + "confusion": [ + [ + 65, + 0, + 5 + ], + [ + 0, + 40, + 0 + ], + [ + 0, + 0, + 40 + ] + ] + }, + "spa": { + "rows": 19, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 19, + "tp": 19, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 19, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + }, + "zh": { + "rows": 186, + "accuracy": 0.8494623655913979, + "macro_f1": 0.8606654783125371, + "macro_f1_present_labels": 0.8606654783125371, + "per_label": { + "AR": { + "support": 82, + "tp": 54, + "fp": 0, + "fn": 28, + "precision": 1.0, + "recall": 0.6585365853658537, + "f1": 0.7941176470588235 + }, + "DIFFUSION": { + "support": 52, + "tp": 52, + "fp": 28, + "fn": 0, + "precision": 0.65, + "recall": 1.0, + "f1": 0.7878787878787878 + }, + "BOTH": { + "support": 52, + "tp": 52, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 54, + 28, + 0 + ], + [ + 0, + 52, + 0 + ], + [ + 0, + 0, + 52 + ] + ] + }, + "zho": { + "rows": 3, + "accuracy": 1.0, + "macro_f1": 0.3333333333333333, + "macro_f1_present_labels": 1.0, + "per_label": { + "AR": { + "support": 3, + "tp": 3, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + }, + "DIFFUSION": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + }, + "BOTH": { + "support": 0, + "tp": 0, + "fp": 0, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0 + } + }, + "confusion": [ + [ + 3, + 0, + 0 + ], + [ + 0, + 0, + 0 + ], + [ + 0, + 0, + 0 + ] + ] + } + }, + "length_bucket": { + "16384": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "32768": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "4096": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "8192": { + "rows": 18, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 6, + "tp": 3, + "fp": 0, + "fn": 3, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 6, + "tp": 6, + "fp": 3, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 6, + "tp": 6, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 3, + 3, + 0 + ], + [ + 0, + 6, + 0 + ], + [ + 0, + 0, + 6 + ] + ] + }, + "short": { + "rows": 929, + "accuracy": 0.8622174381054898, + "macro_f1": 0.8662289383478975, + "macro_f1_present_labels": 0.8662289383478975, + "per_label": { + "AR": { + "support": 449, + "tp": 321, + "fp": 0, + "fn": 128, + "precision": 1.0, + "recall": 0.7149220489977728, + "f1": 0.8337662337662337 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 58, + "fn": 0, + "precision": 0.8053691275167785, + "recall": 1.0, + "f1": 0.8921933085501859 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 70, + "fn": 0, + "precision": 0.7741935483870968, + "recall": 1.0, + "f1": 0.8727272727272727 + } + }, + "confusion": [ + [ + 321, + 58, + 70 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + }, + "position": { + "head": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "middle": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "tail": { + "rows": 24, + "accuracy": 0.8333333333333334, + "macro_f1": 0.8222222222222223, + "macro_f1_present_labels": 0.8222222222222223, + "per_label": { + "AR": { + "support": 8, + "tp": 4, + "fp": 0, + "fn": 4, + "precision": 1.0, + "recall": 0.5, + "f1": 0.6666666666666666 + }, + "DIFFUSION": { + "support": 8, + "tp": 8, + "fp": 4, + "fn": 0, + "precision": 0.6666666666666666, + "recall": 1.0, + "f1": 0.8 + }, + "BOTH": { + "support": 8, + "tp": 8, + "fp": 0, + "fn": 0, + "precision": 1.0, + "recall": 1.0, + "f1": 1.0 + } + }, + "confusion": [ + [ + 4, + 4, + 0 + ], + [ + 0, + 8, + 0 + ], + [ + 0, + 0, + 8 + ] + ] + }, + "unspecified": { + "rows": 929, + "accuracy": 0.8622174381054898, + "macro_f1": 0.8662289383478975, + "macro_f1_present_labels": 0.8662289383478975, + "per_label": { + "AR": { + "support": 449, + "tp": 321, + "fp": 0, + "fn": 128, + "precision": 1.0, + "recall": 0.7149220489977728, + "f1": 0.8337662337662337 + }, + "DIFFUSION": { + "support": 240, + "tp": 240, + "fp": 58, + "fn": 0, + "precision": 0.8053691275167785, + "recall": 1.0, + "f1": 0.8921933085501859 + }, + "BOTH": { + "support": 240, + "tp": 240, + "fp": 70, + "fn": 0, + "precision": 0.7741935483870968, + "recall": 1.0, + "f1": 0.8727272727272727 + } + }, + "confusion": [ + [ + 321, + 58, + 70 + ], + [ + 0, + 240, + 0 + ], + [ + 0, + 0, + 240 + ] + ] + } + } + }, + "actual_tokens": { + "min": 9, + "max": 32768 + }, + "forward_batches_ms": [ + 23.441586992703378, + 21.066898945719004, + 22.36434700898826, + 21.36123296804726, + 21.15163893904537, + 21.08063199557364, + 20.43571590911597, + 20.585373044013977, + 19.55608103889972, + 22.131667006760836, + 20.222938037477434, + 20.64210595563054, + 20.49899601843208, + 21.89586404711008, + 22.36961096059531, + 24.061689036898315, + 21.983468090184033, + 21.385862957686186, + 21.334634977392852, + 21.11936593428254, + 21.973263006657362, + 21.964238956570625, + 21.989220986142755, + 22.06506603397429, + 21.90384001005441, + 21.293926052749157, + 22.208311944268644, + 22.10444596130401, + 22.190683986991644, + 22.173085948452353, + 22.186247049830854, + 22.557210992090404, + 27.591209043748677, + 27.965940069407225, + 24.254580959677696, + 23.27852405142039, + 22.546703927218914, + 22.194349905475974, + 22.648316924460232, + 22.293015965260565, + 22.153826081193984, + 22.296462091617286, + 22.20519504044205, + 22.382849943824112, + 21.579089923761785, + 22.3608179949224, + 22.504118969663978, + 22.45962992310524, + 22.419661981984973, + 22.70016900729388, + 26.77935699466616, + 26.97422809433192, + 23.449325002729893, + 23.269591969437897, + 22.86110504064709, + 23.057644022628665, + 22.436992032453418, + 25.105696986429393, + 23.492714972235262, + 67.84268305636942, + 71.82937895413488, + 64.59801190067083, + 63.158873934298754, + 66.79348403122276, + 61.16222601849586, + 63.127515953965485, + 62.702049035578966, + 63.59631801024079, + 69.06834908295423, + 64.85880399122834, + 65.04428095649928, + 65.86525600869209, + 64.73978504072875, + 65.00474899075925, + 61.46711902692914, + 60.7439239975065, + 61.013582977466285, + 229.41377200186253, + 229.0986479492858, + 240.9975320333615, + 209.29949695710093, + 210.5567289981991, + 205.25914896279573, + 225.68068699911237, + 220.29410605318844, + 229.61657703854144, + 222.52185398247093, + 215.84512200206518, + 207.56465604063123, + 219.98214104678482, + 210.9013630542904, + 221.1646270006895, + 216.87942103017122, + 217.19500294420868, + 214.74841504823416, + 836.3267410313711, + 808.8649999117479, + 836.7812481010333, + 873.6680670408532, + 811.4096500212327, + 841.3394580129534, + 829.4937629252672, + 816.9720850419253, + 823.2158260652795, + 825.6403140258044, + 762.6890699611977, + 810.1025429787114, + 814.1811079112813, + 834.2556380666792, + 875.4450130509213, + 812.7223589690402, + 788.2047790335491, + 864.2736800247803, + 3202.864601975307, + 3210.5051589896902, + 3286.4575510611758, + 3341.1035040626302, + 3360.2551389485598, + 3245.3943740110844, + 3219.6076789405197, + 3393.3652479900047, + 3359.520795987919, + 3324.4257159531116, + 3265.044084051624, + 3187.092025997117, + 3343.4803859563544, + 3285.809724009596, + 3446.741404943168, + 3406.8465729942545, + 3262.78898504097, + 3177.328413934447 + ], + "precision": { + "parameters": "float32", + "forward_autocast": "bfloat16", + "device_type": "cuda" + } +} diff --git a/reproducibility/training-receipts/modality-v6-vela-base/run.json b/reproducibility/training-receipts/modality-v6-vela-base/run.json new file mode 100644 index 0000000000000000000000000000000000000000..7057281d5880134d4b2da12b2b952498484267bf --- /dev/null +++ b/reproducibility/training-receipts/modality-v6-vela-base/run.json @@ -0,0 +1,59 @@ +{ + "base_model": "llm-semantic-router/Vela-1.0-Encoder-307M", + "base_revision": "5fe5bbb1a88b7fdcc93bb5b9d546c574564eb114", + "initial_adapter_sha256": "5d2e619fe94be057a19931ad43122d68e07b34924ff487b4aab89346ff8f4e94", + "train_files": [ + { + "file": "train.jsonl", + "sha256": "33efa6179d7f3954d6dbc08508a8c46627cb36bb72f2f5de14082b1feaf392a1" + }, + { + "file": "train.jsonl", + "sha256": "f51d0566966979cafd62e318da5f1c54acc983672df9ded964c707ec27b3a4d0" + } + ], + "dev_files": [ + { + "file": "dev.jsonl", + "sha256": "bc021f8b5479ab1fd8cfc3cf7e351414dc7e7b429e44b1d0b098d23104aca1d2" + }, + { + "file": "dev.jsonl", + "sha256": "2768c6155a31a5e50c3ef3929fa2bd5ea0a45ea2ed49f4318be3743a69b30b85" + } + ], + "label2id": { + "AR": 0, + "DIFFUSION": 1, + "BOTH": 2 + }, + "train_rows": 19749, + "dev_rows": 1001, + "train_label_counts": { + "AR": 7173, + "DIFFUSION": 6288, + "BOTH": 6288 + }, + "rejected": [], + "seed": 20260913, + "steps": 200, + "batch_size": 1, + "accumulate": 4, + "examples_per_optimizer_step": 4, + "microbatch_token_budget": 32768, + "dropout_trajectory_equivalence": false, + "max_length": 32768, + "learning_rate": 2e-06, + "balanced_sampling": true, + "length_balanced_sampling": true, + "source_balanced_sampling": false, + "source_sampling_probabilities": null, + "implementation_sha256": "2b6a160522dd6efd24893f6146bb641d7f1a8413bd690587b728c028e85d7e3e", + "selection": "source-macro-f1", + "test_used": false, + "precision": "FP32 parameters and explicit FP32 mean CE / BF16 autocast", + "attention": "sdpa", + "checkpointing": "non-reentrant", + "parameters": 314883846, + "trainable_parameters": 7351299 +} diff --git a/reproducibility/training-receipts/modality-v6-vela-base/selection.json b/reproducibility/training-receipts/modality-v6-vela-base/selection.json new file mode 100644 index 0000000000000000000000000000000000000000..17eff397c78ffbccfab2660fdb5559d6dcbfff9b --- /dev/null +++ b/reproducibility/training-receipts/modality-v6-vela-base/selection.json @@ -0,0 +1,6 @@ +{ + "step": 200, + "score": 0.5723292403589801, + "metric": "source-macro-f1", + "test_used": false +} diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..be4ad795f49075ac5e56d8108ef63be314a1e967 --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,55 @@ +{ + "additional_special_tokens": [ + "", + "" + ], + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "cls_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "sep_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/standalone-verification.json b/standalone-verification.json new file mode 100644 index 0000000000000000000000000000000000000000..26ba580b4aa6d2c10d00283b26a7c7852f48de78 --- /dev/null +++ b/standalone-verification.json @@ -0,0 +1,50 @@ +{ + "dtype": "float32", + "provider": "Torch CPU SDPA", + "probability_atol": 0.0002, + "probability_rtol": 0.0001, + "all_argmax_equal_required": true, + "quality_benchmark": false, + "long_context_qualification": false, + "results": [ + { + "batch": 2, + "padded_tokens": 14, + "attention_mask_counts": [ + 14, + 10 + ], + "input_sha256": "5b14566354399d634cd9322d69f32fbe1a3400dcc272481d96318e8370184777", + "maximum_absolute_logit_difference": 1.3828277587890625e-05, + "maximum_absolute_probability_difference": 3.259629011154175e-09, + "argmax_differences": 0, + "passed": true + }, + { + "batch": 2, + "padded_tokens": 20, + "attention_mask_counts": [ + 20, + 11 + ], + "input_sha256": "efd7d799cc51fc1a3a5dacdb84ad5267feb972e6498a7824d1d51e424b2082d2", + "maximum_absolute_logit_difference": 1.6450881958007812e-05, + "maximum_absolute_probability_difference": 2.1141022443771362e-07, + "argmax_differences": 0, + "passed": true + }, + { + "batch": 2, + "padded_tokens": 102, + "attention_mask_counts": [ + 6, + 102 + ], + "input_sha256": "3af336fced079e4eeb53e968b3bd26cbb3ed2c996b37d710c6108708ba5f8770", + "maximum_absolute_logit_difference": 8.463859558105469e-05, + "maximum_absolute_probability_difference": 2.6635825634002686e-06, + "argmax_differences": 0, + "passed": true + } + ] +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..3f81aca7d6f978a836be323226630e754367f1a0 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22fd4a60565f25fee8ebd2866aab07456dd46cbbeb45eaf506529ed8451ee58f +size 34363343 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8b7a2254239805140f894102e5bdc199186057b3 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,2018 @@ +{ + "add_bos_token": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "5": { + "content": "<2mass>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "6": { + "content": "[@BOS@]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "7": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "8": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "9": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "10": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "11": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "12": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "13": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "14": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "15": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "16": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "17": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "18": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "19": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "20": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "21": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "22": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "23": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "24": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "25": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "26": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "27": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "28": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "29": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "30": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "31": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "32": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "33": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "34": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "35": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "36": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "37": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "38": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "39": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "40": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "41": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "42": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "43": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "44": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "45": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "46": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "47": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "48": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "49": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "50": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "51": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "52": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "53": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "54": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "55": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "56": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "57": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "58": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "59": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "60": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "61": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "62": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "63": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "64": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "65": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "66": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "67": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "68": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "69": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "70": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "71": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "72": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "73": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "74": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "75": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "76": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "77": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "78": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "79": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "80": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "81": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "82": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "83": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "84": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "85": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "86": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "87": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "88": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "89": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "90": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "91": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "92": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "93": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "94": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "95": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "96": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "97": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "98": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "99": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "100": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "101": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "102": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "103": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "104": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "105": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "106": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "107": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "108": { + "content": "\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "109": { + "content": "\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "110": { + "content": "\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "111": { + "content": "\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "112": { + "content": "\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "113": { + "content": "\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "114": { + "content": "\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "115": { + "content": "\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "116": { + "content": "\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "117": { + "content": "\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "118": { + "content": "\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "119": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "120": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "121": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "122": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "123": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "124": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "125": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "126": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "127": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "128": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "129": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "130": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "131": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "132": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "133": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "134": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "135": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "136": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "137": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "138": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "139": { + "content": "▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "140": { + "content": "▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "141": { + "content": "▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "142": { + "content": "▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "143": { + "content": "▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "144": { + "content": "▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "145": { + "content": "▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "146": { + "content": "▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "147": { + "content": "▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "148": { + "content": "▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "149": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "150": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "152": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "153": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "154": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "155": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "156": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "157": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "158": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "159": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "160": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "161": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "162": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "163": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "164": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "165": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "166": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "167": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "168": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "169": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "170": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "172": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "173": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "174": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "175": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "171": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "176": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "177": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "178": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "179": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "180": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "181": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "182": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "183": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "184": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "185": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "186": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "187": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "188": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "189": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "190": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "191": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "192": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "193": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "194": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "195": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "196": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "197": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "198": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "199": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "200": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "201": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "202": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "203": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "204": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "205": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "206": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "207": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "208": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "209": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "210": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "211": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "212": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "213": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "214": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "215": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "216": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255968": { + "content": "[toxicity=0]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255969": { + "content": "\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255970": { + "content": "\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255971": { + "content": "\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255972": { + "content": "\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255973": { + "content": "\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255974": { + "content": "\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255975": { + "content": "\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255976": { + "content": "\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255977": { + "content": "\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255978": { + "content": "\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255979": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255980": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255981": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255982": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255983": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255984": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255985": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255986": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255987": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255988": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255989": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255990": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255991": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255992": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255993": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255994": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255995": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255996": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255997": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255998": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255999": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "", + "" + ], + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 32768, + "pad_token": "", + "padding_side": "right", + "sep_token": "", + "spaces_between_special_tokens": false, + "tokenizer_class": "PreTrainedTokenizerFast", + "unk_token": "" +}