Publish MIC usage, slim fixture and reviewed Transformers expectations
Browse files- .mic-repo.json +7 -0
- README.md +37 -0
- build.json +135 -0
- config.json +43 -0
- integration.json +78 -0
- integrations/transformers/expectations/cpu-x86_64-bf16/language-perceiver-mic.bfloat16.json +83 -0
- integrations/transformers/expectations/cpu-x86_64-bf16/language-perceiver.bfloat16.json +83 -0
- integrations/transformers/expectations/cpu-x86_64/language-perceiver-mic.eager.json +83 -0
- integrations/transformers/expectations/cpu-x86_64/language-perceiver-mic.json +83 -0
- integrations/transformers/expectations/cpu-x86_64/language-perceiver.eager.json +83 -0
- integrations/transformers/expectations/cpu-x86_64/language-perceiver.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver-mic.bfloat16.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver-mic.eager.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver-mic.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver.bfloat16.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver.eager.json +83 -0
- integrations/transformers/expectations/cuda/language-perceiver.json +83 -0
- integrations/transformers/usage.md +80 -0
- model.safetensors +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
.mic-repo.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"format": "mic-umbrella-v1",
|
| 3 |
+
"contract_files": [
|
| 4 |
+
"integration.json",
|
| 5 |
+
"integrations/transformers/usage.md"
|
| 6 |
+
]
|
| 7 |
+
}
|
README.md
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- mic
|
| 4 |
+
- test-artifact
|
| 5 |
+
base_model: deepmind/language-perceiver
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# language-perceiver-mic
|
| 9 |
+
|
| 10 |
+
Slim model weights, executable usage and framework expectations in one repository.
|
| 11 |
+
|
| 12 |
+
- [Contract manifest](integration.json)
|
| 13 |
+
- [Transformers usage](integrations/transformers/usage.md)
|
| 14 |
+
- [Transformers expectations](integrations/transformers/expectations)
|
| 15 |
+
- [Fixture provenance](build.json)
|
| 16 |
+
|
| 17 |
+
## Contract
|
| 18 |
+
|
| 19 |
+
### Perceiver masked language modeling
|
| 20 |
+
|
| 21 |
+
Runnable usage and the framework expectations directory are linked above.
|
| 22 |
+
|
| 23 |
+
## Slim model
|
| 24 |
+
|
| 25 |
+
### Slim fixture: language-perceiver-mic
|
| 26 |
+
|
| 27 |
+
A 194.1M-parameter fixture for the [Perceiver MIC](https://huggingface.co/huggingface-mic/language-perceiver-mic), covering [`deepmind/language-perceiver`](https://huggingface.co/deepmind/language-perceiver). It is a test artifact, not a usable model, tagged `test-artifact` to keep it out of candidate-model search.
|
| 28 |
+
|
| 29 |
+
- **Fixture key `de355e4dc541`.** It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
|
| 30 |
+
- **Same architecture, smaller config:** `PerceiverForMaskedLM`, shrunk by the generic rule in [`sweep.py`](https://github.com/huggingface/transformers-mic/blob/main/experiments/sweep.py) ().
|
| 31 |
+
- **Processor files copied byte for byte** from `deepmind/language-perceiver`: `special_tokens_map.json`, `tokenizer_config.json`.
|
| 32 |
+
- **Random weights (seed 17).** They passed the input-sensitivity gate: every perturbation of the contract's inputs moved the logits by at least 0.001 (weakest: `attention_mask:all_ones`, 0.106), so no training was needed.
|
| 33 |
+
- **Same tensor names as the real checkpoint.** The real checkpoint has every parameter, and so does the fixture.
|
| 34 |
+
- **Saved in float32**, the dtype `deepmind/language-perceiver` loads in by default.
|
| 35 |
+
- **Finite in float32, bfloat16, and float16** (forward, loss, gradients).
|
| 36 |
+
|
| 37 |
+
Built by [`build_fixtures.py`](https://github.com/huggingface/transformers-mic/blob/main/example/build_fixtures.py); [`build.json`](build.json) records the pinned sources, gate results, and file hashes. This repository includes the slim weights, usage and framework expectations; rebuild the other files with `python example/build_fixtures.py --fixture language-perceiver-mic`.
|
build.json
ADDED
|
@@ -0,0 +1,135 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"fixture": "huggingface-mic/language-perceiver-mic",
|
| 3 |
+
"key": "de355e4dc541831cc4c2086166054be8ceb0d3b120ebce38703e8abfb7f87829",
|
| 4 |
+
"key_fields": {
|
| 5 |
+
"architectures": [
|
| 6 |
+
"PerceiverForMaskedLM"
|
| 7 |
+
],
|
| 8 |
+
"output_hidden_states": false,
|
| 9 |
+
"return_dict": true,
|
| 10 |
+
"chunk_size_feed_forward": 0,
|
| 11 |
+
"is_encoder_decoder": false,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "LABEL_0",
|
| 14 |
+
"1": "LABEL_1"
|
| 15 |
+
},
|
| 16 |
+
"label2id": {
|
| 17 |
+
"LABEL_0": 0,
|
| 18 |
+
"LABEL_1": 1
|
| 19 |
+
},
|
| 20 |
+
"problem_type": null,
|
| 21 |
+
"num_latents": 256,
|
| 22 |
+
"d_latents": 1280,
|
| 23 |
+
"d_model": 64,
|
| 24 |
+
"num_blocks": 1,
|
| 25 |
+
"num_self_attends_per_block": 26,
|
| 26 |
+
"num_self_attention_heads": 8,
|
| 27 |
+
"num_cross_attention_heads": 8,
|
| 28 |
+
"qk_channels": 256,
|
| 29 |
+
"v_channels": 1280,
|
| 30 |
+
"cross_attention_shape_for_attention": "kv",
|
| 31 |
+
"self_attention_widening_factor": 1,
|
| 32 |
+
"cross_attention_widening_factor": 1,
|
| 33 |
+
"hidden_act": "gelu",
|
| 34 |
+
"attention_probs_dropout_prob": 0.1,
|
| 35 |
+
"initializer_range": 0.02,
|
| 36 |
+
"layer_norm_eps": 1e-12,
|
| 37 |
+
"use_query_residual": true,
|
| 38 |
+
"vocab_size": 262,
|
| 39 |
+
"max_position_embeddings": 2048,
|
| 40 |
+
"image_size": 56,
|
| 41 |
+
"train_size": [
|
| 42 |
+
368,
|
| 43 |
+
496
|
| 44 |
+
],
|
| 45 |
+
"num_frames": 16,
|
| 46 |
+
"audio_samples_per_frame": 1920,
|
| 47 |
+
"samples_per_patch": 16,
|
| 48 |
+
"output_shape": [
|
| 49 |
+
1,
|
| 50 |
+
16,
|
| 51 |
+
224,
|
| 52 |
+
224
|
| 53 |
+
],
|
| 54 |
+
"output_num_channels": 512,
|
| 55 |
+
"_label_trainable_num_channels": 1024,
|
| 56 |
+
"model_type": "perceiver",
|
| 57 |
+
"output_attentions": false
|
| 58 |
+
},
|
| 59 |
+
"covers": {
|
| 60 |
+
"deepmind/language-perceiver": "47a1ba43d8198ea10cf9a158fce114953e9441a1"
|
| 61 |
+
},
|
| 62 |
+
"built_from": {
|
| 63 |
+
"repo": "deepmind/language-perceiver",
|
| 64 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1"
|
| 65 |
+
},
|
| 66 |
+
"seed": 17,
|
| 67 |
+
"dtype": "float32",
|
| 68 |
+
"builder": "example/build_fixtures.py",
|
| 69 |
+
"model_class": "PerceiverForMaskedLM",
|
| 70 |
+
"parameters": 194113286,
|
| 71 |
+
"missing_parameters": [],
|
| 72 |
+
"checkpoint_inspection": "parameter headers/indexes; bounded binary fallback",
|
| 73 |
+
"numeric_health_scope": "CPU float32/bfloat16/float16; autocast for low-precision losses",
|
| 74 |
+
"captured": {
|
| 75 |
+
"call": "training step",
|
| 76 |
+
"inputs": [
|
| 77 |
+
"attention_mask",
|
| 78 |
+
"inputs",
|
| 79 |
+
"labels"
|
| 80 |
+
],
|
| 81 |
+
"class": "PerceiverForMaskedLM"
|
| 82 |
+
},
|
| 83 |
+
"weights": "random",
|
| 84 |
+
"gate": {
|
| 85 |
+
"random": {
|
| 86 |
+
"pass": true,
|
| 87 |
+
"accepted_insensitive": {},
|
| 88 |
+
"threshold": 0.001,
|
| 89 |
+
"deltas": {
|
| 90 |
+
"attention_mask:all_ones": 0.10622552782297134
|
| 91 |
+
},
|
| 92 |
+
"weakest": "attention_mask:all_ones",
|
| 93 |
+
"no_perturbation_defined": [
|
| 94 |
+
"inputs"
|
| 95 |
+
]
|
| 96 |
+
}
|
| 97 |
+
},
|
| 98 |
+
"numeric_health": {
|
| 99 |
+
"float32": {
|
| 100 |
+
"outputs_finite": true,
|
| 101 |
+
"loss_finite": true,
|
| 102 |
+
"gradients_finite": true,
|
| 103 |
+
"finite": true
|
| 104 |
+
},
|
| 105 |
+
"bfloat16": {
|
| 106 |
+
"outputs_finite": true,
|
| 107 |
+
"loss_finite": true,
|
| 108 |
+
"gradients_finite": true,
|
| 109 |
+
"finite": true
|
| 110 |
+
},
|
| 111 |
+
"float16": {
|
| 112 |
+
"outputs_finite": true,
|
| 113 |
+
"loss_finite": true,
|
| 114 |
+
"gradients_finite": true,
|
| 115 |
+
"finite": true
|
| 116 |
+
}
|
| 117 |
+
},
|
| 118 |
+
"copied_unchanged": [
|
| 119 |
+
"special_tokens_map.json",
|
| 120 |
+
"tokenizer_config.json"
|
| 121 |
+
],
|
| 122 |
+
"environment": {
|
| 123 |
+
"python": "3.11.14",
|
| 124 |
+
"machine": "arm64",
|
| 125 |
+
"torch": "2.14.1",
|
| 126 |
+
"transformers": "5.20.1",
|
| 127 |
+
"tokenizers": "0.23.3"
|
| 128 |
+
},
|
| 129 |
+
"sha256": {
|
| 130 |
+
"config.json": "8a00596028a7d53fde5782061e185fdf6a5902d080e440b11e69d99593db06ed",
|
| 131 |
+
"model.safetensors": "da10fb9094185afcaba694447b7e60e0c84b60345c2c13bb477187abc0ec8338",
|
| 132 |
+
"special_tokens_map.json": "08f4e103f6d851b4c14edd06d593dfb25db3339e05299393f08ad53e7733e454",
|
| 133 |
+
"tokenizer_config.json": "1f818797228803a29d7889c3614edfa19d8b3a6742669825df20306ad8ea3abe"
|
| 134 |
+
}
|
| 135 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_label_trainable_num_channels": 1024,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"PerceiverForMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"audio_samples_per_frame": 1920,
|
| 8 |
+
"cross_attention_shape_for_attention": "kv",
|
| 9 |
+
"cross_attention_widening_factor": 1,
|
| 10 |
+
"d_latents": 1280,
|
| 11 |
+
"d_model": 64,
|
| 12 |
+
"dtype": "float32",
|
| 13 |
+
"hidden_act": "gelu",
|
| 14 |
+
"image_size": 56,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"layer_norm_eps": 1e-12,
|
| 17 |
+
"max_position_embeddings": 2048,
|
| 18 |
+
"model_type": "perceiver",
|
| 19 |
+
"num_blocks": 1,
|
| 20 |
+
"num_cross_attention_heads": 8,
|
| 21 |
+
"num_frames": 16,
|
| 22 |
+
"num_latents": 256,
|
| 23 |
+
"num_self_attends_per_block": 26,
|
| 24 |
+
"num_self_attention_heads": 8,
|
| 25 |
+
"output_num_channels": 512,
|
| 26 |
+
"output_shape": [
|
| 27 |
+
1,
|
| 28 |
+
16,
|
| 29 |
+
224,
|
| 30 |
+
224
|
| 31 |
+
],
|
| 32 |
+
"qk_channels": 256,
|
| 33 |
+
"samples_per_patch": 16,
|
| 34 |
+
"self_attention_widening_factor": 1,
|
| 35 |
+
"train_size": [
|
| 36 |
+
368,
|
| 37 |
+
496
|
| 38 |
+
],
|
| 39 |
+
"transformers_version": "5.20.1",
|
| 40 |
+
"use_query_residual": true,
|
| 41 |
+
"v_channels": 1280,
|
| 42 |
+
"vocab_size": 262
|
| 43 |
+
}
|
integration.json
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"mic": "0.3",
|
| 3 |
+
"name": "Perceiver masked language modeling",
|
| 4 |
+
"usage": "PerceiverTokenizer + PerceiverForMaskedLM, byte-level masked LM with two sentences",
|
| 5 |
+
"architecture": "PerceiverForMaskedLM",
|
| 6 |
+
"checkpoints": {
|
| 7 |
+
"deepmind/language-perceiver": {
|
| 8 |
+
"fixture": "huggingface-mic/language-perceiver-mic"
|
| 9 |
+
}
|
| 10 |
+
},
|
| 11 |
+
"fixtures": {
|
| 12 |
+
"huggingface-mic/language-perceiver-mic": {
|
| 13 |
+
"built_from": "deepmind/language-perceiver",
|
| 14 |
+
"covers": [
|
| 15 |
+
"deepmind/language-perceiver"
|
| 16 |
+
],
|
| 17 |
+
"build": {
|
| 18 |
+
"seed": 17
|
| 19 |
+
}
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"integrations": {
|
| 23 |
+
"transformers": {
|
| 24 |
+
"entry_point": "integrations/transformers/usage.md",
|
| 25 |
+
"format": "runnable-markdown",
|
| 26 |
+
"requires": [
|
| 27 |
+
"transformers>=5",
|
| 28 |
+
"torch"
|
| 29 |
+
],
|
| 30 |
+
"targets": [
|
| 31 |
+
"fixtures",
|
| 32 |
+
"checkpoints"
|
| 33 |
+
],
|
| 34 |
+
"variants": {
|
| 35 |
+
"default": {},
|
| 36 |
+
"eager": {
|
| 37 |
+
"attn_implementation": "eager"
|
| 38 |
+
},
|
| 39 |
+
"bfloat16": {
|
| 40 |
+
"dtype": "bfloat16"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"training": true,
|
| 44 |
+
"resources": {
|
| 45 |
+
"fixture": {
|
| 46 |
+
"memory_gb": 1,
|
| 47 |
+
"download_mb": 50
|
| 48 |
+
},
|
| 49 |
+
"deepmind/language-perceiver": {
|
| 50 |
+
"memory_gb": 3,
|
| 51 |
+
"download_mb": 1100,
|
| 52 |
+
"gpu_memory_gb": 2
|
| 53 |
+
}
|
| 54 |
+
}
|
| 55 |
+
}
|
| 56 |
+
},
|
| 57 |
+
"outputs": {
|
| 58 |
+
"runtime": {
|
| 59 |
+
"compare": "exact"
|
| 60 |
+
},
|
| 61 |
+
"masked_logits": {
|
| 62 |
+
"compare": "allclose",
|
| 63 |
+
"atol": 0.0001,
|
| 64 |
+
"rtol": 0.0001
|
| 65 |
+
},
|
| 66 |
+
"training.loss": {
|
| 67 |
+
"compare": "allclose",
|
| 68 |
+
"atol": 0.0001,
|
| 69 |
+
"rtol": 0.0001
|
| 70 |
+
},
|
| 71 |
+
"training.gradients": {
|
| 72 |
+
"compare": "all_true"
|
| 73 |
+
},
|
| 74 |
+
"training.updated": {
|
| 75 |
+
"compare": "true"
|
| 76 |
+
}
|
| 77 |
+
}
|
| 78 |
+
}
|
integrations/transformers/expectations/cpu-x86_64-bf16/language-perceiver-mic.bfloat16.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "bfloat16",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 8.54,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64-bf16"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "bfloat16",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.01190185546875,
|
| 54 |
+
-0.01123046875,
|
| 55 |
+
0.020263671875,
|
| 56 |
+
-0.0120849609375,
|
| 57 |
+
-0.0277099609375
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.0118408203125,
|
| 61 |
+
-0.01123046875,
|
| 62 |
+
0.0201416015625,
|
| 63 |
+
-0.0120849609375,
|
| 64 |
+
-0.027587890625
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.01177978515625,
|
| 68 |
+
-0.01116943359375,
|
| 69 |
+
0.0201416015625,
|
| 70 |
+
-0.0120849609375,
|
| 71 |
+
-0.027587890625
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.5,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cpu-x86_64-bf16/language-perceiver.bfloat16.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "bfloat16",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 8.91,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64-bf16"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "bfloat16",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.75,
|
| 54 |
+
-16.5,
|
| 55 |
+
-16.375,
|
| 56 |
+
-11.8125,
|
| 57 |
+
-16.875
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.1875,
|
| 61 |
+
-12.4375,
|
| 62 |
+
-12.25,
|
| 63 |
+
-15.3125,
|
| 64 |
+
-12.25
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.75,
|
| 68 |
+
-13.0,
|
| 69 |
+
-12.8125,
|
| 70 |
+
-16.25,
|
| 71 |
+
-12.75
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.6875,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cpu-x86_64/language-perceiver-mic.eager.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "eager",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 10.64,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.011857529170811176,
|
| 54 |
+
-0.011009469628334045,
|
| 55 |
+
0.020300013944506645,
|
| 56 |
+
-0.012026689946651459,
|
| 57 |
+
-0.027553845196962357
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.011856728233397007,
|
| 61 |
+
-0.011009469628334045,
|
| 62 |
+
0.02030036970973015,
|
| 63 |
+
-0.012030145153403282,
|
| 64 |
+
-0.027552612125873566
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.011856435798108578,
|
| 68 |
+
-0.011008954606950283,
|
| 69 |
+
0.020306482911109924,
|
| 70 |
+
-0.01202866155654192,
|
| 71 |
+
-0.02755468711256981
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.569273471832275,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cpu-x86_64/language-perceiver-mic.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "default",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 10.52,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.011857529170811176,
|
| 54 |
+
-0.011009469628334045,
|
| 55 |
+
0.020300013944506645,
|
| 56 |
+
-0.012026689946651459,
|
| 57 |
+
-0.027553845196962357
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.011856728233397007,
|
| 61 |
+
-0.011009469628334045,
|
| 62 |
+
0.02030036970973015,
|
| 63 |
+
-0.012030145153403282,
|
| 64 |
+
-0.027552612125873566
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.011856435798108578,
|
| 68 |
+
-0.011008954606950283,
|
| 69 |
+
0.020306482911109924,
|
| 70 |
+
-0.01202866155654192,
|
| 71 |
+
-0.02755468711256981
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.569273471832275,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cpu-x86_64/language-perceiver.eager.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "eager",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 10.64,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.573013305664062,
|
| 54 |
+
-16.39983367919922,
|
| 55 |
+
-16.173620223999023,
|
| 56 |
+
-11.96962833404541,
|
| 57 |
+
-16.72991943359375
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.123590469360352,
|
| 61 |
+
-12.390275001525879,
|
| 62 |
+
-12.191612243652344,
|
| 63 |
+
-15.236956596374512,
|
| 64 |
+
-12.192231178283691
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.757885932922363,
|
| 68 |
+
-12.936495780944824,
|
| 69 |
+
-12.775068283081055,
|
| 70 |
+
-16.288785934448242,
|
| 71 |
+
-12.742189407348633
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.6892828941345215,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cpu-x86_64/language-perceiver.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "default",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.17",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cpu",
|
| 15 |
+
"torchvision": "0.29.1+cpu",
|
| 16 |
+
"torchaudio": "2.11.0+cpu",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.3.5",
|
| 19 |
+
"pillow": "12.3.0",
|
| 20 |
+
"accelerate": "1.15.0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 10.96,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cpu (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cpu (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.3.5 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "12.3.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.15.0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cpu-x86_64"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cpu",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.573013305664062,
|
| 54 |
+
-16.39983367919922,
|
| 55 |
+
-16.173620223999023,
|
| 56 |
+
-11.96962833404541,
|
| 57 |
+
-16.72991943359375
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.123590469360352,
|
| 61 |
+
-12.390275001525879,
|
| 62 |
+
-12.191612243652344,
|
| 63 |
+
-15.236956596374512,
|
| 64 |
+
-12.192231178283691
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.757885932922363,
|
| 68 |
+
-12.936495780944824,
|
| 69 |
+
-12.775068283081055,
|
| 70 |
+
-16.288785934448242,
|
| 71 |
+
-12.742189407348633
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.6892828941345215,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver-mic.bfloat16.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "bfloat16",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.39,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "bfloat16",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.0118408203125,
|
| 54 |
+
-0.01129150390625,
|
| 55 |
+
0.020263671875,
|
| 56 |
+
-0.01202392578125,
|
| 57 |
+
-0.0277099609375
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.0118408203125,
|
| 61 |
+
-0.01129150390625,
|
| 62 |
+
0.020263671875,
|
| 63 |
+
-0.01202392578125,
|
| 64 |
+
-0.0277099609375
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.0118408203125,
|
| 68 |
+
-0.0113525390625,
|
| 69 |
+
0.0203857421875,
|
| 70 |
+
-0.011962890625,
|
| 71 |
+
-0.0277099609375
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.5625,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver-mic.eager.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "eager",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.62,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.011857532896101475,
|
| 54 |
+
-0.01100946869701147,
|
| 55 |
+
0.02030000649392605,
|
| 56 |
+
-0.012026693671941757,
|
| 57 |
+
-0.027553843334317207
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.011856730096042156,
|
| 61 |
+
-0.01100947242230177,
|
| 62 |
+
0.02030036970973015,
|
| 63 |
+
-0.012030155397951603,
|
| 64 |
+
-0.027552619576454163
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.011856437660753727,
|
| 68 |
+
-0.011008958332240582,
|
| 69 |
+
0.02030649036169052,
|
| 70 |
+
-0.012028663419187069,
|
| 71 |
+
-0.027554688975214958
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.569441318511963,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver-mic.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "default",
|
| 8 |
+
"target": "language-perceiver-mic",
|
| 9 |
+
"kind": "fixture",
|
| 10 |
+
"revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.51,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
0.011857532896101475,
|
| 54 |
+
-0.01100946869701147,
|
| 55 |
+
0.02030000649392605,
|
| 56 |
+
-0.012026693671941757,
|
| 57 |
+
-0.027553843334317207
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
0.011856730096042156,
|
| 61 |
+
-0.01100947242230177,
|
| 62 |
+
0.02030036970973015,
|
| 63 |
+
-0.012030155397951603,
|
| 64 |
+
-0.027552619576454163
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
0.011856437660753727,
|
| 68 |
+
-0.011008958332240582,
|
| 69 |
+
0.02030649036169052,
|
| 70 |
+
-0.012028663419187069,
|
| 71 |
+
-0.027554688975214958
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 5.569441318511963,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver.bfloat16.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "bfloat16",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.27,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "bfloat16",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.75,
|
| 54 |
+
-16.5,
|
| 55 |
+
-16.25,
|
| 56 |
+
-11.8125,
|
| 57 |
+
-16.875
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.1875,
|
| 61 |
+
-12.4375,
|
| 62 |
+
-12.25,
|
| 63 |
+
-15.25,
|
| 64 |
+
-12.25
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.75,
|
| 68 |
+
-12.9375,
|
| 69 |
+
-12.8125,
|
| 70 |
+
-16.375,
|
| 71 |
+
-12.75
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.8125,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver.eager.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "eager",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.29,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.573028564453125,
|
| 54 |
+
-16.399845123291016,
|
| 55 |
+
-16.173633575439453,
|
| 56 |
+
-11.969634056091309,
|
| 57 |
+
-16.729930877685547
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.123595237731934,
|
| 61 |
+
-12.390280723571777,
|
| 62 |
+
-12.191617012023926,
|
| 63 |
+
-15.236940383911133,
|
| 64 |
+
-12.19223690032959
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.757892608642578,
|
| 68 |
+
-12.936500549316406,
|
| 69 |
+
-12.775073051452637,
|
| 70 |
+
-16.288787841796875,
|
| 71 |
+
-12.742195129394531
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.059415817260742,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/expectations/cuda/language-perceiver.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"recorded_with": {
|
| 3 |
+
"contract": "huggingface-mic/language-perceiver-mic",
|
| 4 |
+
"contract_name": "Perceiver",
|
| 5 |
+
"contract_revision": "103ee7b7c5faff0bb516dc2550021d3e5453241a",
|
| 6 |
+
"framework": "transformers",
|
| 7 |
+
"variant": "default",
|
| 8 |
+
"target": "language-perceiver",
|
| 9 |
+
"kind": "checkpoint",
|
| 10 |
+
"revision": "47a1ba43d8198ea10cf9a158fce114953e9441a1",
|
| 11 |
+
"environment": {
|
| 12 |
+
"python": "3.11.18",
|
| 13 |
+
"transformers": "5.21.0.dev0",
|
| 14 |
+
"torch": "2.14.0+cu130",
|
| 15 |
+
"torchvision": "0.29.1+cu130",
|
| 16 |
+
"torchaudio": "2.11.0+cu130",
|
| 17 |
+
"tokenizers": "0.23.3",
|
| 18 |
+
"numpy": "2.4.6",
|
| 19 |
+
"pillow": "13.0.0",
|
| 20 |
+
"accelerate": "1.16.0.dev0",
|
| 21 |
+
"safetensors": "0.8.0",
|
| 22 |
+
"huggingface-hub": "2.3.0",
|
| 23 |
+
"sentencepiece": "0.2.2",
|
| 24 |
+
"protobuf": "7.36.2",
|
| 25 |
+
"librosa": "0.11.0",
|
| 26 |
+
"soundfile": "0.14.0",
|
| 27 |
+
"scipy": "1.17.1"
|
| 28 |
+
},
|
| 29 |
+
"seconds": 7.32,
|
| 30 |
+
"environment_differs": {
|
| 31 |
+
"torch": "2.14.0+cu130 (pinned 2.11.0)",
|
| 32 |
+
"torchvision": "0.29.1+cu130 (pinned 0.26.0)",
|
| 33 |
+
"tokenizers": "0.23.3 (pinned 0.23.1)",
|
| 34 |
+
"numpy": "2.4.6 (pinned 1.26.4)",
|
| 35 |
+
"pillow": "13.0.0 (pinned 12.2.0)",
|
| 36 |
+
"accelerate": "1.16.0.dev0 (pinned 1.13.0)",
|
| 37 |
+
"huggingface-hub": "2.3.0 (pinned 1.32.0)",
|
| 38 |
+
"sentencepiece": "0.2.2 (pinned 0.2.1)",
|
| 39 |
+
"protobuf": "7.36.2 (pinned 6.33.6)",
|
| 40 |
+
"soundfile": "0.14.0 (pinned 0.13.1)",
|
| 41 |
+
"scipy": "1.17.1 (pinned 1.15.3)"
|
| 42 |
+
},
|
| 43 |
+
"platform": "cuda"
|
| 44 |
+
},
|
| 45 |
+
"outputs": {
|
| 46 |
+
"runtime": {
|
| 47 |
+
"device": "cuda",
|
| 48 |
+
"dtype": "float32",
|
| 49 |
+
"attn_implementation": "eager"
|
| 50 |
+
},
|
| 51 |
+
"masked_logits": [
|
| 52 |
+
[
|
| 53 |
+
-16.573028564453125,
|
| 54 |
+
-16.399845123291016,
|
| 55 |
+
-16.173633575439453,
|
| 56 |
+
-11.969634056091309,
|
| 57 |
+
-16.729930877685547
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
-12.123595237731934,
|
| 61 |
+
-12.390280723571777,
|
| 62 |
+
-12.191617012023926,
|
| 63 |
+
-15.236940383911133,
|
| 64 |
+
-12.19223690032959
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
-12.757892608642578,
|
| 68 |
+
-12.936500549316406,
|
| 69 |
+
-12.775073051452637,
|
| 70 |
+
-16.288787841796875,
|
| 71 |
+
-12.742195129394531
|
| 72 |
+
]
|
| 73 |
+
],
|
| 74 |
+
"training": {
|
| 75 |
+
"loss": 6.059415817260742,
|
| 76 |
+
"gradients": {
|
| 77 |
+
"perceiver.input_preprocessor.embeddings.weight": true,
|
| 78 |
+
"perceiver.encoder.self_attends.0.attention.self.query.weight": true
|
| 79 |
+
},
|
| 80 |
+
"updated": true
|
| 81 |
+
}
|
| 82 |
+
}
|
| 83 |
+
}
|
integrations/transformers/usage.md
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Perceiver: masked language modeling with Transformers
|
| 2 |
+
|
| 3 |
+
This page is the Transformers contract of the Perceiver MIC and also a usage example. [doc-builder](https://github.com/huggingface/doc-builder/blob/main/docs/runnable-code-blocks.md) renders the `runnable` blocks below as documentation, and its pytest plugin runs them as written against `deepmind/language-perceiver`. The MIC runner runs the same blocks against each slim fixture and each checkpoint in [`integration.json`](../../integration.json). Lines marked `# doc-builder: hide` only matter there: they swap in the pinned checkpoint or variant and print the result record.
|
| 4 |
+
|
| 5 |
+
## Predict the masked bytes
|
| 6 |
+
|
| 7 |
+
```py runnable:masked_lm
|
| 8 |
+
import torch
|
| 9 |
+
from transformers import PerceiverTokenizer, PerceiverForMaskedLM, set_seed
|
| 10 |
+
import json # doc-builder: hide
|
| 11 |
+
import os # doc-builder: hide
|
| 12 |
+
|
| 13 |
+
model_id = "deepmind/language-perceiver"
|
| 14 |
+
model_id = os.environ.get("MIC_MODEL", model_id) # doc-builder: hide
|
| 15 |
+
variant = json.loads(os.environ.get("MIC_MODEL_KWARGS", "{}")) # doc-builder: hide
|
| 16 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 17 |
+
set_seed(0)
|
| 18 |
+
|
| 19 |
+
tokenizer = PerceiverTokenizer.from_pretrained(model_id)
|
| 20 |
+
model = PerceiverForMaskedLM.from_pretrained(model_id).to(device)
|
| 21 |
+
if variant: # doc-builder: hide
|
| 22 |
+
del model # doc-builder: hide
|
| 23 |
+
model = PerceiverForMaskedLM.from_pretrained(model_id, **variant).to(device) # doc-builder: hide
|
| 24 |
+
|
| 25 |
+
# Two short sentences; the tokenizer maps each character to its byte value.
|
| 26 |
+
# The decoder predicts a fixed number of positions, so pad to that length.
|
| 27 |
+
texts = ["The cat sat on the mat.", "A dog ran in the park."]
|
| 28 |
+
encoding = tokenizer(texts, padding="max_length", max_length=model.config.max_position_embeddings, return_tensors="pt")
|
| 29 |
+
input_ids = encoding["input_ids"].to(device)
|
| 30 |
+
attention_mask = encoding["attention_mask"].to(device)
|
| 31 |
+
|
| 32 |
+
original_ids = input_ids.clone()
|
| 33 |
+
|
| 34 |
+
# Mask "mat" in the first sentence: characters 19-21, shifted by one for the leading [CLS] token
|
| 35 |
+
start = texts[0].index("mat") + 1
|
| 36 |
+
input_ids[0, start : start + 3] = tokenizer.mask_token_id
|
| 37 |
+
|
| 38 |
+
with torch.no_grad():
|
| 39 |
+
outputs = model(inputs=input_ids, attention_mask=attention_mask)
|
| 40 |
+
logits = outputs.logits
|
| 41 |
+
|
| 42 |
+
# Compact probe: first 5 logits at each masked position
|
| 43 |
+
masked = input_ids == tokenizer.mask_token_id
|
| 44 |
+
masked_logits = logits[masked]
|
| 45 |
+
print(masked_logits[:, :5].tolist())
|
| 46 |
+
```
|
| 47 |
+
|
| 48 |
+
## One training step
|
| 49 |
+
|
| 50 |
+
```py runnable:masked_lm:2
|
| 51 |
+
# Labels: the original bytes; ignore pad positions with -100
|
| 52 |
+
labels = original_ids.clone()
|
| 53 |
+
labels[attention_mask == 0] = -100
|
| 54 |
+
|
| 55 |
+
model.train()
|
| 56 |
+
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)
|
| 57 |
+
loss = model(inputs=input_ids, attention_mask=attention_mask, labels=labels).loss
|
| 58 |
+
loss.backward()
|
| 59 |
+
before = {n: p.detach().cpu().clone() for n, p in model.named_parameters() if p.requires_grad} # doc-builder: hide
|
| 60 |
+
optimizer.step()
|
| 61 |
+
print(f"loss: {loss.item():.4f}")
|
| 62 |
+
|
| 63 |
+
if os.environ.get("MIC_RUNNER"): # doc-builder: hide
|
| 64 |
+
named = dict(model.named_parameters())
|
| 65 |
+
checked = ("perceiver.input_preprocessor.embeddings.weight", "perceiver.encoder.self_attends.0.attention.self.query.weight")
|
| 66 |
+
result = {
|
| 67 |
+
"runtime": {
|
| 68 |
+
"device": model.device.type,
|
| 69 |
+
"dtype": str(model.dtype).removeprefix("torch."),
|
| 70 |
+
"attn_implementation": model.config._attn_implementation,
|
| 71 |
+
},
|
| 72 |
+
"masked_logits": masked_logits[:, :5].float().tolist(),
|
| 73 |
+
"training": {
|
| 74 |
+
"loss": loss.item(),
|
| 75 |
+
"gradients": {n: bool(named[n].grad is not None and torch.count_nonzero(named[n].grad) > 0 and torch.isfinite(named[n].grad).all()) for n in checked},
|
| 76 |
+
"updated": any(not torch.equal(p.detach().cpu(), before[n]) for n, p in model.named_parameters() if n in before),
|
| 77 |
+
},
|
| 78 |
+
}
|
| 79 |
+
print(json.dumps(result))
|
| 80 |
+
```
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da10fb9094185afcaba694447b7e60e0c84b60345c2c13bb477187abc0ec8338
|
| 3 |
+
size 776511992
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": {"content": "[BOS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "eos_token": {"content": "[EOS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "sep_token": {"content": "[SEP]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": {"content": "[PAD]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "cls_token": {"content": "[CLS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "mask_token": {"content": "[MASK]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"pad_token": {"content": "[PAD]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "bos_token": {"content": "[BOS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "[EOS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "mask_token": {"content": "[MASK]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "cls_token": {"content": "[CLS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "sep_token": {"content": "[SEP]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 2048, "tokenizer_class": "PerceiverTokenizer"}
|