pod backup 2026-11-07: e33 (0-66)

e33/S1-E33-M10-rt-n16k-a100-syn/DONE

@@ -0,0 +1 @@
+2026-10-24 12:31:59

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter/README.md

@@ -0,0 +1,206 @@
+---
+base_model: /workspace/jev_kt/models/hf/JEV-9B
+library_name: peft
+tags:
+- base_model:adapter:/workspace/jev_kt/models/hf/JEV-9B
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.21.2

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter/adapter_config.json

@@ -0,0 +1,57 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": "/workspace/jev_kt/models/hf/JEV-9B",
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "q_proj",
+    "out_proj",
+    "down_proj",
+    "v_proj",
+    "o_proj",
+    "k_proj",
+    "up_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a8ca54346d8be0f2b16b0d745cf749492f4d53dd9e2d475c96efee98b01839
+size 160486456

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_config.json

@@ -0,0 +1,58 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": null,
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64.0,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "k_proj",
+    "lm_head",
+    "o_proj",
+    "out_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:923bb59225a7e6a362da7f25ac4849ad5380fd7e8caba63ef739f292bec44675
+size 176641832

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.00045958356349729,
+  -0.0004598347295541316,
+  -0.0022042542695999146,
+  -0.0013978549977764487,
+  -0.00023735652212053537,
+  0.0020917123183608055,
+  0.0016224675346165895,
+  -0.002466864651069045,
+  -0.0005318059120327234,
+  -8.45411341288127e-05,
+  -7.737032137811184e-05,
+  0.0007597437361255288,
+  -0.00017794813902582973,
+  -0.0037094249855726957,
+  -0.004070935305207968,
+  -0.005084027070552111,
+  -0.005918621551245451,
+  -0.004334166180342436,
+  -0.004935281816869974,
+  -0.005004563368856907,
+  -0.00402710260823369,
+  -0.003410839010030031,
+  -0.005615165922790766,
+  -0.0034875874407589436
+ ],
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "note": "decision logits = lm_head-LoRA logprobs of verbalizer ids + bias, then per-kind temperature"
+}

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.7724464809625557,
+  "choice": 1.0270244461420386,
+  "score": 0.9929109453941127
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.7724464809625557,
+   "ce_before": 0.5699275732040405,
+   "ce_after": 0.5652568340301514
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0270244461420386,
+   "ce_before": 0.6716148257255554,
+   "ce_after": 0.6714505553245544
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9929109453941127,
+   "ce_before": 1.0511720180511475,
+   "ce_after": 1.0511550903320312
+  }
+ },
+ "source": "models/e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000"
+}

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9330c8f5460cbe3894d3cf5eccd9ceca662a9daaeb7a787d139aec96c6c1990b
+size 393464

e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000/judge_config.json

@@ -0,0 +1,164 @@
+{
+ "base_model_path": "/root/models/Qwen3.5-9B",
+ "hidden_size": 4096,
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "kinds": [
+  "noul",
+  "choice",
+  "score"
+ ],
+ "model_name": "jev-judge-qwen35-9b",
+ "model_version": "0.8.0",
+ "lm_head_included": true,
+ "weights_mode": "unmerged",
+ "adapter_subfolder": "adapter",
+ "lora": {
+  "r": 16,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "target_modules": [
+   "down_proj",
+   "gate_proj",
+   "in_proj_qkv",
+   "in_proj_z",
+   "k_proj",
+   "o_proj",
+   "out_proj",
+   "q_proj",
+   "up_proj",
+   "v_proj"
+  ]
+ },
+ "trained_stage": "e16",
+ "source_checkpoint": "models/e33/S1-E33-M10-rt-n16k-a100-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e33/E33-M10-rt-n16k-a100",
+  "data_manifest": {
+   "arm": "E33-M10-rt-n16k-a100",
+   "name": null,
+   "train_sha256": "e0cc7b84a6deb613bff644587c8fc7dda043c2c3266043dc7df4c0c79180428f",
+   "n_train": 16000
+  },
+  "init": "released",
+  "objective": "s-log",
+  "sigma": 0.5,
+  "K": 8,
+  "reward": "brier",
+  "clip": 0.2,
+  "passes": 1,
+  "lr": 0.0001,
+  "lr_head": 0.0002,
+  "rows_per_step": 32,
+  "train_seed": 0,
+  "replay": "data/e16/replay/train.jsonl",
+  "replay_frac": 0.05,
+  "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e",
+  "max_len": 3072,
+  "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps",
+  "total_rows": 16000,
+  "lora": {
+   "r": 16,
+   "lora_alpha": 32,
+   "lora_dropout": 0.05
+  },
+  "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3",
+  "event": "ckpt",
+  "tag": "ckpt-16000",
+  "rows_seen": 16000,
+  "replay_seen": 842,
+  "step": 500,
+  "tokens": 9820167,
+  "tok_per_s": 854.7992102709214,
+  "rows_per_s": 1.4660166465301607,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5699275732040405,
+   "noul_T": 0.7724464809625557,
+   "choice_ce": 0.6716148257255554,
+   "choice_T": 1.0270244461420386,
+   "score_ce": 1.0511720180511475,
+   "score_T": 0.9929109453941127,
+   "noul_auc": 0.7252821333333332,
+   "noul_acc": 0.699,
+   "noul_f1": 0.7762081784386617,
+   "noul_nll": 0.5881762043386698,
+   "noul_ece": 0.059616081506013885,
+   "noul_brier": 0.201311856030733,
+   "noul_sharp": 0.1597384661436081,
+   "entropy": 0.6134923100471497
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.236003875732422
+ }
+}

e33/S1-E33-M10-rt-n16k-a100-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "09:20:27", "event": "start", "data": "data/e33/E33-M10-rt-n16k-a100", "data_manifest": {"arm": "E33-M10-rt-n16k-a100", "name": null, "train_sha256": "e0cc7b84a6deb613bff644587c8fc7dda043c2c3266043dc7df4c0c79180428f", "n_train": 16000}, "init": "released", "objective": "s-log", "sigma": 0.5, "K": 8, "reward": "brier", "clip": 0.2, "passes": 1, "lr": 0.0001, "lr_head": 0.0002, "rows_per_step": 32, "train_seed": 0, "replay": "data/e16/replay/train.jsonl", "replay_frac": 0.05, "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e", "max_len": 3072, "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps", "total_rows": 16000, "lora": {"r": 16, "lora_alpha": 32, "lora_dropout": 0.05}, "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3", "device": "cuda", "load_s": 77.06579637527466}
+{"time": "09:24:08", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 891.0474855226341, "loss": 0.6033337190747261, "entropy": 0.6253835633397102, "sharp": 0.15371828898787498}
+{"time": "09:27:49", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 898.7678657186418, "loss": 0.7838748916983604, "entropy": 0.6202669739723206, "sharp": 0.17797065526247025}
+{"time": "09:31:29", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 900.5382753446562, "loss": 0.6850737277418375, "entropy": 0.6641247570514679, "sharp": 0.10811960324645042}
+{"time": "09:35:01", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 900.3117851157423, "loss": 0.6847868859767914, "entropy": 0.6819761171936989, "sharp": 0.05960580799728632}
+{"time": "09:38:33", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 900.4366446851287, "loss": 0.5863783843815327, "entropy": 0.575633630156517, "sharp": 0.19002662599086761}
+{"time": "09:42:12", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 903.3374492922703, "loss": 0.5031997989863157, "entropy": 0.6125824078917503, "sharp": 0.1816745474934578}
+{"time": "09:45:44", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 903.7665703758253, "loss": 0.606873907148838, "entropy": 0.6636044085025787, "sharp": 0.10620175953954458}
+{"time": "09:49:21", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 903.3837596631068, "loss": 0.67482590675354, "entropy": 0.6118676699697971, "sharp": 0.15935452934354544}
+{"time": "09:52:53", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 903.3855343113702, "loss": 0.6431846879422665, "entropy": 0.6385258324444294, "sharp": 0.1421925025060773}
+{"time": "09:56:36", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 903.9834459122642, "loss": 0.6846701428294182, "entropy": 0.6390752457082272, "sharp": 0.14440846256911755}
+{"time": "09:58:00", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.6084082722663879, "noul_T": 0.6766352911501362, "noul_auc": 0.7320551551872544, "noul_acc": 0.702, "noul_f1": 0.764612954186414, "noul_nll": 0.6007609467208386, "noul_ece": 0.10155431270599366, "noul_brier": 0.20541022797599395, "noul_sharp": 0.13225561380386353, "entropy": 0.6460034251213074}
+{"time": "10:01:37", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 873.0366049535218, "loss": 0.6482396014034748, "entropy": 0.6171336583793163, "sharp": 0.17463437281548977}
+{"time": "10:05:05", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 875.1641802576061, "loss": 0.5996145531535149, "entropy": 0.6567100770771503, "sharp": 0.11905446369200945}
+{"time": "10:08:38", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 876.9032903879504, "loss": 0.6384086310863495, "entropy": 0.6788572743535042, "sharp": 0.06982394959777594}
+{"time": "10:12:14", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 878.9804242819296, "loss": 0.5768000446259975, "entropy": 0.6693370379507542, "sharp": 0.09189106337726116}
+{"time": "10:15:49", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 880.4348184722512, "loss": 0.6017400855198503, "entropy": 0.5698162391781807, "sharp": 0.21178777143359184}
+{"time": "10:19:27", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 881.5823790589051, "loss": 0.6542864292860031, "entropy": 0.5234885523095727, "sharp": 0.2540337862446904}
+{"time": "10:23:00", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 882.7998236958621, "loss": 0.5910687409341335, "entropy": 0.6635227352380753, "sharp": 0.10423886124044657}
+{"time": "10:26:37", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 883.9654856469632, "loss": 0.5948001351207495, "entropy": 0.6536477170884609, "sharp": 0.11783802974969149}
+{"time": "10:30:13", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 884.6143325550358, "loss": 0.6801256858743727, "entropy": 0.6189041864126921, "sharp": 0.16625779680907726}
+{"time": "10:33:53", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 885.7659601500735, "loss": 0.6023488454520702, "entropy": 0.6618353426456451, "sharp": 0.10348301753401756}
+{"time": "10:35:16", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.613629937171936, "noul_T": 0.8413533877688976, "noul_auc": 0.7025704207523923, "noul_acc": 0.674, "noul_f1": 0.7634252539912917, "noul_nll": 0.6114704060256482, "noul_ece": 0.06213674157857896, "noul_brier": 0.21065328342820616, "noul_sharp": 0.14712855219841003, "entropy": 0.6329501271247864}
+{"time": "10:38:58", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 871.2871728674021, "loss": 0.63688157312572, "entropy": 0.6147606391459703, "sharp": 0.17704907432198524}
+{"time": "10:42:32", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 872.5228054587427, "loss": 0.6519083455204964, "entropy": 0.6725699678063393, "sharp": 0.08100294321775436}
+{"time": "10:46:13", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 873.2950804448907, "loss": 0.6356181167066097, "entropy": 0.6462880596518517, "sharp": 0.12306220084428787}
+{"time": "10:49:53", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 874.3815137635282, "loss": 0.6321443058550358, "entropy": 0.652955025434494, "sharp": 0.11930426955223083}
+{"time": "10:53:33", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 875.7611935040712, "loss": 0.6394157260656357, "entropy": 0.6025685258209705, "sharp": 0.1883924026042223}
+{"time": "10:57:14", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 876.6742822955374, "loss": 0.6191202476620674, "entropy": 0.6671462059020996, "sharp": 0.09179499931633472}
+{"time": "11:00:47", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 877.4646161094252, "loss": 0.6553642265498638, "entropy": 0.6592379845678806, "sharp": 0.1110451053828001}
+{"time": "11:04:26", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 877.971221797522, "loss": 0.6024574413895607, "entropy": 0.6117802187800407, "sharp": 0.17410529032349586}
+{"time": "11:07:55", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 878.6396164404216, "loss": 0.6674257572740316, "entropy": 0.6629147864878178, "sharp": 0.10723049286752939}
+{"time": "11:11:39", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 879.5878966609004, "loss": 0.6327837016433477, "entropy": 0.6372476555407047, "sharp": 0.14154105819761753}
+{"time": "11:13:02", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.5927555561065674, "noul_T": 0.9140837438194751, "noul_auc": 0.7317513628461967, "noul_acc": 0.666, "noul_f1": 0.7727891156462585, "noul_nll": 0.5982471793740988, "noul_ece": 0.07837080037593842, "noul_brier": 0.20586142497615276, "noul_sharp": 0.17416174709796906, "entropy": 0.6058613657951355}
+{"time": "11:16:37", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 869.6137309110122, "loss": 0.6072483845055103, "entropy": 0.5658436417579651, "sharp": 0.20860496163368225}
+{"time": "11:20:14", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 870.7963165762047, "loss": 0.6785305291414261, "entropy": 0.6227179393172264, "sharp": 0.1529350159689784}
+{"time": "11:23:52", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 871.4461092863057, "loss": 0.6080680526793003, "entropy": 0.640902929008007, "sharp": 0.1289919689297676}
+{"time": "11:27:28", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 872.4537510502763, "loss": 0.6423982172273099, "entropy": 0.6232390608638525, "sharp": 0.1561298780143261}
+{"time": "11:31:08", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 873.4323797092313, "loss": 0.65891507640481, "entropy": 0.6288746707141399, "sharp": 0.14588055666536093}
+{"time": "11:34:44", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 874.1338999630548, "loss": 0.6176988501101732, "entropy": 0.6252961456775665, "sharp": 0.13359309546649456}
+{"time": "11:38:28", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 874.6831089882023, "loss": 0.744150634855032, "entropy": 0.6299264468252659, "sharp": 0.1376659180969}
+{"time": "11:41:58", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 875.1158409723324, "loss": 0.5802575796842575, "entropy": 0.6352507062256336, "sharp": 0.14148460514843464}
+{"time": "11:45:39", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 875.5994456569073, "loss": 0.7059405953623354, "entropy": 0.6364332772791386, "sharp": 0.12399807758629322}
+{"time": "11:49:25", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 876.3615689884818, "loss": 0.6372781246900558, "entropy": 0.6148921549320221, "sharp": 0.15839910320937634}
+{"time": "11:50:48", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5831224322319031, "noul_T": 0.973220551345713, "noul_auc": 0.7322745607669069, "noul_acc": 0.71, "noul_f1": 0.7792998477929984, "noul_nll": 0.5908164388295263, "noul_ece": 0.06788681399822236, "noul_brier": 0.20089749663804382, "noul_sharp": 0.18074163794517517, "entropy": 0.5872051119804382}
+{"time": "11:54:30", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 869.1485172518481, "loss": 0.5657886452972889, "entropy": 0.6486924085766077, "sharp": 0.11286498792469501}
+{"time": "11:58:14", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 870.0233991919794, "loss": 0.5447256825864315, "entropy": 0.6085926666855812, "sharp": 0.16909427661448717}
+{"time": "12:02:00", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 870.6947731484165, "loss": 0.6258899383246899, "entropy": 0.6019481364637613, "sharp": 0.16560775693506002}
+{"time": "12:05:35", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 871.4700596618806, "loss": 0.6144727133214474, "entropy": 0.6315025687217712, "sharp": 0.1404762426391244}
+{"time": "12:09:21", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 872.1645453687298, "loss": 0.5760183185338974, "entropy": 0.656073272228241, "sharp": 0.10922098997980356}
+{"time": "12:12:58", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 872.7241558853473, "loss": 0.6531402003020048, "entropy": 0.6523221842944622, "sharp": 0.12087107170373201}
+{"time": "12:16:37", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 873.3651392286066, "loss": 0.6043087020516396, "entropy": 0.6481751687824726, "sharp": 0.1281145429238677}
+{"time": "12:20:16", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 873.6749996879207, "loss": 0.6318537872284651, "entropy": 0.6360546965152025, "sharp": 0.14198052696883678}
+{"time": "12:23:56", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 874.2588882022271, "loss": 0.6480370759963989, "entropy": 0.6473933756351471, "sharp": 0.11683804169297218}
+{"time": "12:27:35", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 874.6643402276964, "loss": 0.6334035154432058, "entropy": 0.6499393992125988, "sharp": 0.12684703432023525}
+{"time": "12:28:58", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5850250124931335, "noul_T": 0.7595249686054862, "noul_auc": 0.7318188722553206, "noul_acc": 0.702, "noul_f1": 0.7682737169517885, "noul_nll": 0.590603579364717, "noul_ece": 0.06252238416671752, "noul_brier": 0.20186487573155232, "noul_sharp": 0.1538901925086975, "entropy": 0.6182107925415039}
+{"time": "12:31:59", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9820167, "tok_per_s": 854.7992102709214, "rows_per_s": 1.4660166465301607, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5699275732040405, "noul_T": 0.7724464809625557, "choice_ce": 0.6716148257255554, "choice_T": 1.0270244461420386, "score_ce": 1.0511720180511475, "score_T": 0.9929109453941127, "noul_auc": 0.7252821333333332, "noul_acc": 0.699, "noul_f1": 0.7762081784386617, "noul_nll": 0.5881762043386698, "noul_ece": 0.059616081506013885, "noul_brier": 0.201311856030733, "noul_sharp": 0.1597384661436081, "entropy": 0.6134923100471497}, "dropped_long": 0, "peak_gib": 18.236003875732422}

e33/S1-E33-M10-rt-n16k-syn/DONE

@@ -0,0 +1 @@
+2026-10-23 04:32:35

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter/README.md

@@ -0,0 +1,206 @@
+---
+base_model: /workspace/jev_kt/models/hf/JEV-9B
+library_name: peft
+tags:
+- base_model:adapter:/workspace/jev_kt/models/hf/JEV-9B
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.21.2

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter/adapter_config.json

@@ -0,0 +1,57 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": "/workspace/jev_kt/models/hf/JEV-9B",
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "out_proj",
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "up_proj",
+    "in_proj_qkv",
+    "gate_proj",
+    "in_proj_z",
+    "down_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acd7398bcc7b780b9076cd2cd710473219c65da73c87a462f1bf7db725e9bcd1
+size 160486456

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter_vllm/adapter_config.json

@@ -0,0 +1,58 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": null,
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64.0,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "k_proj",
+    "lm_head",
+    "o_proj",
+    "out_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1296a823eceecc39bff4c46d0e1947b9302822acce5e88ddf53fcd4389962d5c
+size 176641832

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.0003609308914747089,
+  -0.0003611863648984581,
+  -0.002497137524187565,
+  -0.0015482216840609908,
+  -0.0005122810835018754,
+  0.002160529140383005,
+  0.0014962729765102267,
+  -0.00209335470572114,
+  -0.0005815276526845992,
+  -7.023397483862936e-05,
+  2.4424463845207356e-05,
+  0.0003453893295954913,
+  -0.00025366476620547473,
+  -0.003527699038386345,
+  -0.004392900038510561,
+  -0.005109368823468685,
+  -0.00589381018653512,
+  -0.004366089124232531,
+  -0.004935386590659618,
+  -0.00498735997825861,
+  -0.003816308919340372,
+  -0.0033886691089719534,
+  -0.005703995935618877,
+  -0.0034793075174093246
+ ],
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "note": "decision logits = lm_head-LoRA logprobs of verbalizer ids + bias, then per-kind temperature"
+}

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.7800532877395329,
+  "choice": 1.0466806238139477,
+  "score": 0.9804041633946682
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.7800532877395329,
+   "ce_before": 0.5706181526184082,
+   "ce_after": 0.5660994648933411
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0466806238139477,
+   "ce_before": 0.672325849533081,
+   "ce_after": 0.6718399524688721
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9804041633946682,
+   "ce_before": 1.0526014566421509,
+   "ce_after": 1.0524842739105225
+  }
+ },
+ "source": "models/e33/S1-E33-M10-rt-n16k-syn/ckpt-16000"
+}

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1ea418ef3e6efca8f1a8bd124c42306c9abf5f03aae0d9dec4a89cf24f60687
+size 393464

e33/S1-E33-M10-rt-n16k-syn/ckpt-16000/judge_config.json

@@ -0,0 +1,164 @@
+{
+ "base_model_path": "/root/models/Qwen3.5-9B",
+ "hidden_size": 4096,
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "kinds": [
+  "noul",
+  "choice",
+  "score"
+ ],
+ "model_name": "jev-judge-qwen35-9b",
+ "model_version": "0.8.0",
+ "lm_head_included": true,
+ "weights_mode": "unmerged",
+ "adapter_subfolder": "adapter",
+ "lora": {
+  "r": 16,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "target_modules": [
+   "down_proj",
+   "gate_proj",
+   "in_proj_qkv",
+   "in_proj_z",
+   "k_proj",
+   "o_proj",
+   "out_proj",
+   "q_proj",
+   "up_proj",
+   "v_proj"
+  ]
+ },
+ "trained_stage": "e16",
+ "source_checkpoint": "models/e33/S1-E33-M10-rt-n16k-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e33/E33-M10-rt-n16k",
+  "data_manifest": {
+   "arm": "E33-M10-rt-n16k",
+   "name": null,
+   "train_sha256": "39574ddaa158a99572227f29fcc2be761352b45f2f70805ff1c7883379ef2228",
+   "n_train": 16000
+  },
+  "init": "released",
+  "objective": "s-log",
+  "sigma": 0.5,
+  "K": 8,
+  "reward": "brier",
+  "clip": 0.2,
+  "passes": 1,
+  "lr": 0.0001,
+  "lr_head": 0.0002,
+  "rows_per_step": 32,
+  "train_seed": 0,
+  "replay": "data/e16/replay/train.jsonl",
+  "replay_frac": 0.05,
+  "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e",
+  "max_len": 3072,
+  "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps",
+  "total_rows": 16000,
+  "lora": {
+   "r": 16,
+   "lora_alpha": 32,
+   "lora_dropout": 0.05
+  },
+  "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3",
+  "event": "ckpt",
+  "tag": "ckpt-16000",
+  "rows_seen": 16000,
+  "replay_seen": 842,
+  "step": 500,
+  "tokens": 9820167,
+  "tok_per_s": 853.3575821037343,
+  "rows_per_s": 1.4635441932658964,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5706181526184082,
+   "noul_T": 0.7800532877395329,
+   "choice_ce": 0.672325849533081,
+   "choice_T": 1.0466806238139477,
+   "score_ce": 1.0526014566421509,
+   "score_T": 0.9804041633946682,
+   "noul_auc": 0.7283541333333334,
+   "noul_acc": 0.698,
+   "noul_f1": 0.7719033232628398,
+   "noul_nll": 0.5866083004325628,
+   "noul_ece": 0.05212497982382775,
+   "noul_brier": 0.20057906316333285,
+   "noul_sharp": 0.16233906149864197,
+   "entropy": 0.6147943139076233
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.236003875732422
+ }
+}

e33/S1-E33-M10-rt-n16k-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "01:20:46", "event": "start", "data": "data/e33/E33-M10-rt-n16k", "data_manifest": {"arm": "E33-M10-rt-n16k", "name": null, "train_sha256": "39574ddaa158a99572227f29fcc2be761352b45f2f70805ff1c7883379ef2228", "n_train": 16000}, "init": "released", "objective": "s-log", "sigma": 0.5, "K": 8, "reward": "brier", "clip": 0.2, "passes": 1, "lr": 0.0001, "lr_head": 0.0002, "rows_per_step": 32, "train_seed": 0, "replay": "data/e16/replay/train.jsonl", "replay_frac": 0.05, "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e", "max_len": 3072, "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps", "total_rows": 16000, "lora": {"r": 16, "lora_alpha": 32, "lora_dropout": 0.05}, "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3", "device": "cuda", "load_s": 68.17974424362183}
+{"time": "01:24:26", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 894.3302368416892, "loss": 0.6407141238451004, "entropy": 0.6530866995453835, "sharp": 0.11467453930526972}
+{"time": "01:28:07", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 898.5712368978669, "loss": 0.7670875117182732, "entropy": 0.5850043445825577, "sharp": 0.19216507486999035}
+{"time": "01:31:47", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 901.0519607560497, "loss": 0.7097546458244324, "entropy": 0.5784373320639133, "sharp": 0.2102995691820979}
+{"time": "01:35:19", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 900.2845654414082, "loss": 0.7051415480673313, "entropy": 0.6509120985865593, "sharp": 0.11466245353221893}
+{"time": "01:38:52", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 899.3461954764921, "loss": 0.6417021974921227, "entropy": 0.6230399906635284, "sharp": 0.16287371329963207}
+{"time": "01:42:32", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 902.1724212711766, "loss": 0.5933823697268963, "entropy": 0.6530337370932102, "sharp": 0.1106806006282568}
+{"time": "01:46:04", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 902.7435101228933, "loss": 0.6372644007205963, "entropy": 0.6816029846668243, "sharp": 0.060833889059722424}
+{"time": "01:49:43", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 901.7272742006631, "loss": 0.6699827425181866, "entropy": 0.6287437509745359, "sharp": 0.1442428957670927}
+{"time": "01:53:15", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 901.3173038745224, "loss": 0.6240650974214077, "entropy": 0.6412920206785202, "sharp": 0.1320671895518899}
+{"time": "01:57:00", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 901.9174655351277, "loss": 0.6517882402986288, "entropy": 0.6239370983093977, "sharp": 0.15501127019524574}
+{"time": "01:58:23", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.5938156843185425, "noul_T": 0.7507550395879644, "noul_auc": 0.7372365023375133, "noul_acc": 0.714, "noul_f1": 0.7906295754026355, "noul_nll": 0.5917655223309993, "noul_ece": 0.06725453257560729, "noul_brier": 0.20199009536325635, "noul_sharp": 0.14850011467933655, "entropy": 0.6278250813484192}
+{"time": "02:02:00", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 871.1336387513961, "loss": 0.6488936170935631, "entropy": 0.6459421142935753, "sharp": 0.1318453261628747}
+{"time": "02:05:30", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 873.0455970204106, "loss": 0.6333391219377518, "entropy": 0.6602405309677124, "sharp": 0.0976364342495799}
+{"time": "02:09:03", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 874.8895332796241, "loss": 0.646759383380413, "entropy": 0.5998613610863686, "sharp": 0.1769361738115549}
+{"time": "02:12:40", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 876.7330429188311, "loss": 0.6052324529737234, "entropy": 0.6510478630661964, "sharp": 0.10768228489905596}
+{"time": "02:16:15", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 878.2464503690718, "loss": 0.6153158433735371, "entropy": 0.631971400231123, "sharp": 0.16027738526463509}
+{"time": "02:19:55", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 879.2137809500738, "loss": 0.6790948957204819, "entropy": 0.5736175198107958, "sharp": 0.20291051920503378}
+{"time": "02:23:29", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 880.3711935175188, "loss": 0.6004719063639641, "entropy": 0.6468684375286102, "sharp": 0.12351896334439516}
+{"time": "02:27:06", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 881.5599301165362, "loss": 0.6054106764495373, "entropy": 0.6608152016997337, "sharp": 0.09726034570485353}
+{"time": "02:30:42", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 882.3999166719803, "loss": 0.6535965269431472, "entropy": 0.6553077027201653, "sharp": 0.11877382826060057}
+{"time": "02:34:22", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 883.6472831287725, "loss": 0.61347995698452, "entropy": 0.6301144063472748, "sharp": 0.14622782915830612}
+{"time": "02:35:45", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.5859304070472717, "noul_T": 0.8799948384607998, "noul_auc": 0.7436330188520024, "noul_acc": 0.716, "noul_f1": 0.7874251497005988, "noul_nll": 0.5828873897492886, "noul_ece": 0.06448433697223664, "noul_brier": 0.1981293654491963, "noul_sharp": 0.17154835164546967, "entropy": 0.6047180891036987}
+{"time": "02:39:28", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 869.194350995947, "loss": 0.6664544409140944, "entropy": 0.6718409713357687, "sharp": 0.08500322513282299}
+{"time": "02:43:03", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 870.3098890608844, "loss": 0.6444327458739281, "entropy": 0.6366416960954666, "sharp": 0.13889538496732712}
+{"time": "02:46:45", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 870.9679598953579, "loss": 0.6266806535422802, "entropy": 0.6547847539186478, "sharp": 0.11867003049701452}
+{"time": "02:50:25", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 872.162954408354, "loss": 0.6558815725147724, "entropy": 0.6340029835700989, "sharp": 0.1507103629410267}
+{"time": "02:54:06", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 873.4662758411903, "loss": 0.642772326245904, "entropy": 0.6260318588465452, "sharp": 0.16632072255015373}
+{"time": "02:57:47", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 874.4971117879062, "loss": 0.6335619688034058, "entropy": 0.6518338844180107, "sharp": 0.12628033384680748}
+{"time": "03:01:20", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 875.2975710934581, "loss": 0.6495114751160145, "entropy": 0.6683188453316689, "sharp": 0.09399524331092834}
+{"time": "03:04:59", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 875.8806219727835, "loss": 0.6487135104835033, "entropy": 0.6416224911808968, "sharp": 0.1189706027507782}
+{"time": "03:08:29", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 876.5671341788939, "loss": 0.6583934891968966, "entropy": 0.6552927903831005, "sharp": 0.11414727661758661}
+{"time": "03:12:12", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 877.6184190910565, "loss": 0.661140576004982, "entropy": 0.6403373926877975, "sharp": 0.13451462611556053}
+{"time": "03:13:36", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.5856049656867981, "noul_T": 0.8816176327090242, "noul_auc": 0.7410507839530135, "noul_acc": 0.684, "noul_f1": 0.7799442896935933, "noul_nll": 0.5889199020117521, "noul_ece": 0.07037900412082672, "noul_brier": 0.20162840001933632, "noul_sharp": 0.1706424504518509, "entropy": 0.6040006875991821}
+{"time": "03:17:11", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 867.636449941811, "loss": 0.6064267233014107, "entropy": 0.6226922497153282, "sharp": 0.1504072267562151}
+{"time": "03:20:48", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 868.825550660966, "loss": 0.6666115932166576, "entropy": 0.6551536694169044, "sharp": 0.10939735546708107}
+{"time": "03:24:27", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 869.444203220318, "loss": 0.6336435861885548, "entropy": 0.6463650986552238, "sharp": 0.12483089882880449}
+{"time": "03:28:03", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 870.5032601087813, "loss": 0.6576888412237167, "entropy": 0.6181784169748425, "sharp": 0.1659212587401271}
+{"time": "03:31:44", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 871.4820855569737, "loss": 0.6604277305305004, "entropy": 0.6469684541225433, "sharp": 0.11548988427966833}
+{"time": "03:35:20", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 872.2455562457085, "loss": 0.6381114348769188, "entropy": 0.6450099423527718, "sharp": 0.11503359023481607}
+{"time": "03:39:04", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 872.8070292594366, "loss": 0.6908784322440624, "entropy": 0.633655920624733, "sharp": 0.13484611455351114}
+{"time": "03:42:33", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 873.3877540759133, "loss": 0.6250544004142284, "entropy": 0.6358201876282692, "sharp": 0.1346720615401864}
+{"time": "03:46:14", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 873.9114803828065, "loss": 0.6878696195781231, "entropy": 0.6477202232927084, "sharp": 0.10971695929765701}
+{"time": "03:50:00", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 874.7192303517428, "loss": 0.6515332609415054, "entropy": 0.6342602297663689, "sharp": 0.13561679888516665}
+{"time": "03:51:23", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5833331346511841, "noul_T": 0.8217515025225614, "noul_auc": 0.740055020168436, "noul_acc": 0.714, "noul_f1": 0.7816793893129771, "noul_nll": 0.584851800262928, "noul_ece": 0.05926298546791077, "noul_brier": 0.1992804793556761, "noul_sharp": 0.16393373906612396, "entropy": 0.6101114749908447}
+{"time": "03:55:05", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 867.5312933374554, "loss": 0.627738818526268, "entropy": 0.6642075832933187, "sharp": 0.09371273778378963}
+{"time": "03:58:49", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 868.4505163717379, "loss": 0.5989300683140755, "entropy": 0.6216349601745605, "sharp": 0.15461522340774536}
+{"time": "04:02:35", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 869.1675793578265, "loss": 0.6381424367427826, "entropy": 0.6140604615211487, "sharp": 0.15680895280092955}
+{"time": "04:06:12", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 869.8846409524637, "loss": 0.6422123610973358, "entropy": 0.6449132598936558, "sharp": 0.12053444609045982}
+{"time": "04:09:58", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 870.5906451027971, "loss": 0.6214579604566097, "entropy": 0.6666248366236687, "sharp": 0.09311869088560343}
+{"time": "04:13:35", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 871.2019719938318, "loss": 0.6460805404931307, "entropy": 0.6567777544260025, "sharp": 0.1107736174017191}
+{"time": "04:17:13", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 871.8909668828694, "loss": 0.6471943072974682, "entropy": 0.6520841270685196, "sharp": 0.12355391774326563}
+{"time": "04:20:53", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 872.1641297327224, "loss": 0.6499506458640099, "entropy": 0.6403382867574692, "sharp": 0.13805944379419088}
+{"time": "04:24:34", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 872.7285434720358, "loss": 0.6431107819080353, "entropy": 0.6519273668527603, "sharp": 0.11090059950947762}
+{"time": "04:28:13", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 873.1511200997285, "loss": 0.6419705078005791, "entropy": 0.6479619462043047, "sharp": 0.12581786420196295}
+{"time": "04:29:36", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5854170918464661, "noul_T": 0.7670265124397351, "noul_auc": 0.7415402271691616, "noul_acc": 0.712, "noul_f1": 0.7714285714285715, "noul_nll": 0.5860456532686948, "noul_ece": 0.062291146516799935, "noul_brier": 0.19968873407987509, "noul_sharp": 0.15745879709720612, "entropy": 0.6191316246986389}
+{"time": "04:32:35", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9820167, "tok_per_s": 853.3575821037343, "rows_per_s": 1.4635441932658964, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5706181526184082, "noul_T": 0.7800532877395329, "choice_ce": 0.672325849533081, "choice_T": 1.0466806238139477, "score_ce": 1.0526014566421509, "score_T": 0.9804041633946682, "noul_auc": 0.7283541333333334, "noul_acc": 0.698, "noul_f1": 0.7719033232628398, "noul_nll": 0.5866083004325628, "noul_ece": 0.05212497982382775, "noul_brier": 0.20057906316333285, "noul_sharp": 0.16233906149864197, "entropy": 0.6147943139076233}, "dropped_long": 0, "peak_gib": 18.236003875732422}

e33/S1-E33-M20P40-n16k-a100-syn/DONE

@@ -0,0 +1 @@
+2026-10-23 17:35:55

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter/README.md

@@ -0,0 +1,206 @@
+---
+base_model: /workspace/jev_kt/models/hf/JEV-9B
+library_name: peft
+tags:
+- base_model:adapter:/workspace/jev_kt/models/hf/JEV-9B
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.21.2

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter/adapter_config.json

@@ -0,0 +1,57 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": "/workspace/jev_kt/models/hf/JEV-9B",
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "out_proj",
+    "in_proj_z",
+    "down_proj",
+    "q_proj",
+    "v_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "up_proj",
+    "k_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b06f968a6db3848efc7e61de7ca5803a555d400c261bfb55f4e605bd2cc15e8
+size 160486456

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_config.json

@@ -0,0 +1,58 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": null,
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64.0,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "k_proj",
+    "lm_head",
+    "o_proj",
+    "out_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:343b9dba12d72646e7c3cad503764c38b9b9bda2d18b929dffd0341d2715f4dc
+size 176641832

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.00010125698463525623,
+  -0.00010150979505851865,
+  -0.002252081409096718,
+  -0.0015649446286261082,
+  -0.0010697965044528246,
+  0.0015692494343966246,
+  0.0011667461367323995,
+  -0.0025911687407642603,
+  -1.5943171092658304e-05,
+  -0.0004627322778105736,
+  0.0002803638344630599,
+  0.0005846124258823693,
+  -0.0008904358255676925,
+  -0.0034926794469356537,
+  -0.0042582787573337555,
+  -0.005071217194199562,
+  -0.005935357417911291,
+  -0.004361479543149471,
+  -0.004931907169520855,
+  -0.004948125220835209,
+  -0.00379412854090333,
+  -0.0033585415221750736,
+  -0.005670496728271246,
+  -0.0034690620377659798
+ ],
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "note": "decision logits = lm_head-LoRA logprobs of verbalizer ids + bias, then per-kind temperature"
+}

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.8002108628973789,
+  "choice": 1.0728412258162334,
+  "score": 0.9491235278105069
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.8002108628973789,
+   "ce_before": 0.5691902041435242,
+   "ce_after": 0.5654259920120239
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0728412258162334,
+   "ce_before": 0.6705418825149536,
+   "ce_after": 0.6693586111068726
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9491235278105069,
+   "ce_before": 1.0455740690231323,
+   "ce_after": 1.0447719097137451
+  }
+ },
+ "source": "models/e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000"
+}

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e7dd06ef2275896db2a9ad8e41f09f0ef7babb8b31997efa4d1e3e013489b25
+size 393464

e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000/judge_config.json

@@ -0,0 +1,164 @@
+{
+ "base_model_path": "/root/models/Qwen3.5-9B",
+ "hidden_size": 4096,
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "kinds": [
+  "noul",
+  "choice",
+  "score"
+ ],
+ "model_name": "jev-judge-qwen35-9b",
+ "model_version": "0.8.0",
+ "lm_head_included": true,
+ "weights_mode": "unmerged",
+ "adapter_subfolder": "adapter",
+ "lora": {
+  "r": 16,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "target_modules": [
+   "down_proj",
+   "gate_proj",
+   "in_proj_qkv",
+   "in_proj_z",
+   "k_proj",
+   "o_proj",
+   "out_proj",
+   "q_proj",
+   "up_proj",
+   "v_proj"
+  ]
+ },
+ "trained_stage": "e16",
+ "source_checkpoint": "models/e33/S1-E33-M20P40-n16k-a100-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e33/E33-M20P40-n16k-a100",
+  "data_manifest": {
+   "arm": "E33-M20P40-n16k-a100",
+   "name": null,
+   "train_sha256": "4c2cbc7e61a30f31b9bd5413d5605555b3ac9485e7e3cb9e113abd1aa1394b12",
+   "n_train": 16000
+  },
+  "init": "released",
+  "objective": "s-log",
+  "sigma": 0.5,
+  "K": 8,
+  "reward": "brier",
+  "clip": 0.2,
+  "passes": 1,
+  "lr": 0.0001,
+  "lr_head": 0.0002,
+  "rows_per_step": 32,
+  "train_seed": 0,
+  "replay": "data/e16/replay/train.jsonl",
+  "replay_frac": 0.05,
+  "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e",
+  "max_len": 3072,
+  "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps",
+  "total_rows": 16000,
+  "lora": {
+   "r": 16,
+   "lora_alpha": 32,
+   "lora_dropout": 0.05
+  },
+  "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3",
+  "event": "ckpt",
+  "tag": "ckpt-16000",
+  "rows_seen": 16000,
+  "replay_seen": 842,
+  "step": 500,
+  "tokens": 9814579,
+  "tok_per_s": 806.9924313028692,
+  "rows_per_s": 1.384814012135178,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5691902041435242,
+   "noul_T": 0.8002108628973789,
+   "choice_ce": 0.6705418825149536,
+   "choice_T": 1.0728412258162334,
+   "score_ce": 1.0455740690231323,
+   "score_T": 0.9491235278105069,
+   "noul_auc": 0.7222741333333333,
+   "noul_acc": 0.698,
+   "noul_f1": 0.7739520958083832,
+   "noul_nll": 0.5880829917564988,
+   "noul_ece": 0.04662516665458681,
+   "noul_brier": 0.20118704083871727,
+   "noul_sharp": 0.1676119565963745,
+   "entropy": 0.6107596158981323
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.236003875732422
+ }
+}

e33/S1-E33-M20P40-n16k-a100-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "14:13:08", "event": "start", "data": "data/e33/E33-M20P40-n16k-a100", "data_manifest": {"arm": "E33-M20P40-n16k-a100", "name": null, "train_sha256": "4c2cbc7e61a30f31b9bd5413d5605555b3ac9485e7e3cb9e113abd1aa1394b12", "n_train": 16000}, "init": "released", "objective": "s-log", "sigma": 0.5, "K": 8, "reward": "brier", "clip": 0.2, "passes": 1, "lr": 0.0001, "lr_head": 0.0002, "rows_per_step": 32, "train_seed": 0, "replay": "data/e16/replay/train.jsonl", "replay_frac": 0.05, "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e", "max_len": 3072, "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps", "total_rows": 16000, "lora": {"r": 16, "lora_alpha": 32, "lora_dropout": 0.05}, "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3", "device": "cuda", "load_s": 120.29126191139221}
+{"time": "14:16:54", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 839.8715478391055, "loss": 0.6187404282391071, "entropy": 0.6414551772177219, "sharp": 0.13162385113537312}
+{"time": "14:20:44", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 847.5550024837847, "loss": 0.710713678970933, "entropy": 0.6639010868966579, "sharp": 0.094074378721416}
+{"time": "14:24:42", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 846.8465414505208, "loss": 0.7149962484836578, "entropy": 0.6894855201244354, "sharp": 0.038734118454158306}
+{"time": "14:28:40", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 847.0194710535048, "loss": 0.7016126066446304, "entropy": 0.6360101252794266, "sharp": 0.15717380121350288}
+{"time": "14:32:21", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 846.5476430796652, "loss": 0.7025396898388863, "entropy": 0.6633415594696999, "sharp": 0.09588960744440556}
+{"time": "14:36:18", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 847.856212094996, "loss": 0.6576686762273312, "entropy": 0.5767230838537216, "sharp": 0.2077706828713417}
+{"time": "14:40:04", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 849.3255868206792, "loss": 0.7632252350449562, "entropy": 0.6031612306833267, "sharp": 0.1766234626993537}
+{"time": "14:43:56", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 848.8644693577511, "loss": 0.562406275421381, "entropy": 0.5192938018590212, "sharp": 0.2664580028504133}
+{"time": "14:47:44", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 847.7852483871987, "loss": 0.677379759028554, "entropy": 0.6627754978835583, "sharp": 0.10070306714624166}
+{"time": "14:51:39", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 847.9684956914746, "loss": 0.6650274395942688, "entropy": 0.6410431265830994, "sharp": 0.14800265431404114}
+{"time": "14:53:08", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.6212870478630066, "noul_T": 0.7023261478585969, "noul_auc": 0.6843260029366593, "noul_acc": 0.672, "noul_f1": 0.7759562841530054, "noul_nll": 0.6250243791341782, "noul_ece": 0.11322395074367522, "noul_brier": 0.21671475337394724, "noul_sharp": 0.12332789599895477, "entropy": 0.6514794230461121}
+{"time": "14:56:48", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 818.6033910951888, "loss": 0.6246248260140419, "entropy": 0.604729138314724, "sharp": 0.19723656587302685}
+{"time": "15:00:34", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 820.7401742662969, "loss": 0.7470029867254198, "entropy": 0.5473670903593302, "sharp": 0.24039341136813164}
+{"time": "15:04:22", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 822.128574168893, "loss": 0.6280730292201042, "entropy": 0.6206810176372528, "sharp": 0.16349924262613058}
+{"time": "15:08:05", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 823.5858737424818, "loss": 0.6877431944012642, "entropy": 0.6785900443792343, "sharp": 0.07349966000765562}
+{"time": "15:11:57", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 825.5068292890887, "loss": 0.7086129002273083, "entropy": 0.6279721036553383, "sharp": 0.14614077657461166}
+{"time": "15:15:50", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 827.0023994783346, "loss": 0.7685284856706858, "entropy": 0.6500500217080116, "sharp": 0.1163442237302661}
+{"time": "15:19:40", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 828.4397478955465, "loss": 0.754140667617321, "entropy": 0.6031205579638481, "sharp": 0.1930545661598444}
+{"time": "15:23:33", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 829.5254864991339, "loss": 0.6428478360176086, "entropy": 0.6708729267120361, "sharp": 0.08859098423272371}
+{"time": "15:27:20", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 830.864614360502, "loss": 0.5738391559571028, "entropy": 0.6217141449451447, "sharp": 0.15805615298449993}
+{"time": "15:31:14", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 831.7710687381879, "loss": 0.6900174766778946, "entropy": 0.6694989874958992, "sharp": 0.0823324890807271}
+{"time": "15:32:43", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.6128379106521606, "noul_T": 0.47963188571230614, "noul_auc": 0.7305193161296856, "noul_acc": 0.7, "noul_f1": 0.7572815533980582, "noul_nll": 0.6143106191158295, "noul_ece": 0.1173224647641182, "noul_brier": 0.21178017520018558, "noul_sharp": 0.10002126544713974, "entropy": 0.6633838415145874}
+{"time": "15:36:35", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 818.3244384803786, "loss": 0.6299922671169043, "entropy": 0.6626281999051571, "sharp": 0.10336934588849545}
+{"time": "15:40:22", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 819.8278409393922, "loss": 0.627080044709146, "entropy": 0.6587131060659885, "sharp": 0.11120752524584532}
+{"time": "15:44:19", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 820.8217398785365, "loss": 0.5289444080553949, "entropy": 0.6182603435590863, "sharp": 0.16458167880773544}
+{"time": "15:48:16", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 821.7753717182602, "loss": 0.668922908604145, "entropy": 0.5956772342324257, "sharp": 0.19893265888094902}
+{"time": "15:52:13", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 822.6517675023122, "loss": 0.6773939281702042, "entropy": 0.6701566651463509, "sharp": 0.08436406124383211}
+{"time": "15:55:57", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 823.7591338250224, "loss": 0.6251176791265607, "entropy": 0.6630509197711945, "sharp": 0.10360329411923885}
+{"time": "15:59:46", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 825.0068217712911, "loss": 0.698808241635561, "entropy": 0.5815245062112808, "sharp": 0.21828553080558777}
+{"time": "16:03:35", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 826.2237881602903, "loss": 0.7049896717071533, "entropy": 0.654325544834137, "sharp": 0.09251557663083076}
+{"time": "16:07:26", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 827.1292792775448, "loss": 0.688546247780323, "entropy": 0.6717874817550182, "sharp": 0.08574923686683178}
+{"time": "16:11:15", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 827.914890733714, "loss": 0.49946608766913414, "entropy": 0.5843364968895912, "sharp": 0.20784184522926807}
+{"time": "16:12:44", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.605241060256958, "noul_T": 1.2048853041749858, "noul_auc": 0.7253042142748646, "noul_acc": 0.636, "noul_f1": 0.7672634271099744, "noul_nll": 0.6115475974157453, "noul_ece": 0.13372548961639405, "noul_brier": 0.21162126575926385, "noul_sharp": 0.2091650366783142, "entropy": 0.5707269906997681}
+{"time": "16:16:32", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 818.8158304626734, "loss": 0.5467460304498672, "entropy": 0.5993260815739632, "sharp": 0.18954907543957233}
+{"time": "16:20:26", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 820.0792956300551, "loss": 0.6160920187830925, "entropy": 0.6452987492084503, "sharp": 0.13316628709435463}
+{"time": "16:24:14", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 821.1997374291061, "loss": 0.5811580680310726, "entropy": 0.6088825017213821, "sharp": 0.16268229950219393}
+{"time": "16:28:05", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 821.9665063572719, "loss": 0.6118759140372276, "entropy": 0.6484171822667122, "sharp": 0.12860039994120598}
+{"time": "16:31:58", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 822.97676320785, "loss": 0.7243249639868736, "entropy": 0.6476022303104401, "sharp": 0.1273578805848956}
+{"time": "16:35:44", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 823.7959610682416, "loss": 0.6667574434541166, "entropy": 0.6401418969035149, "sharp": 0.11859891004860401}
+{"time": "16:39:31", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 824.6030474215052, "loss": 0.6770539954304695, "entropy": 0.6548423916101456, "sharp": 0.11251895409077406}
+{"time": "16:43:23", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 825.3579685095436, "loss": 0.683322167955339, "entropy": 0.6304250992834568, "sharp": 0.15285257250070572}
+{"time": "16:47:17", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 825.9141744084313, "loss": 0.5915361074730754, "entropy": 0.6303008385002613, "sharp": 0.1420151386409998}
+{"time": "16:51:08", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 826.6632978600414, "loss": 0.5277743395417929, "entropy": 0.637439914047718, "sharp": 0.13507275190204382}
+{"time": "16:52:37", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5858119130134583, "noul_T": 0.7325064249704112, "noul_auc": 0.7344348618588715, "noul_acc": 0.702, "noul_f1": 0.7732115677321156, "noul_nll": 0.5899210872650147, "noul_ece": 0.06367846584320071, "noul_brier": 0.20153352968871333, "noul_sharp": 0.14930002391338348, "entropy": 0.6234035491943359}
+{"time": "16:56:25", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 819.9214470641089, "loss": 0.589809961616993, "entropy": 0.6252590082585812, "sharp": 0.14650190621614456}
+{"time": "17:00:14", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 820.7593532438311, "loss": 0.5925923585891724, "entropy": 0.6398867517709732, "sharp": 0.11890413705259562}
+{"time": "17:04:05", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 821.6220787132297, "loss": 0.6462609097361565, "entropy": 0.6125414073467255, "sharp": 0.16461621969938278}
+{"time": "17:07:57", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 822.3114539874771, "loss": 0.6334951892495155, "entropy": 0.6172429770231247, "sharp": 0.15678761340677738}
+{"time": "17:11:47", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 823.0525671531353, "loss": 0.7070290148258209, "entropy": 0.6357797607779503, "sharp": 0.1409710068255663}
+{"time": "17:15:34", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 823.6646668729059, "loss": 0.6082311794161797, "entropy": 0.6516525633633137, "sharp": 0.1177465096116066}
+{"time": "17:19:30", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 824.2039480742526, "loss": 0.633880028501153, "entropy": 0.6386780217289925, "sharp": 0.13634685520082712}
+{"time": "17:23:27", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 824.8682535596918, "loss": 0.5927850492298603, "entropy": 0.6280360221862793, "sharp": 0.1544718537479639}
+{"time": "17:27:15", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 825.4752101102926, "loss": 0.6072592251002789, "entropy": 0.6435235626995564, "sharp": 0.12449884880334139}
+{"time": "17:31:12", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 825.8721817850699, "loss": 0.5903935097157955, "entropy": 0.6376286670565605, "sharp": 0.14049519877880812}
+{"time": "17:32:41", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5839874148368835, "noul_T": 0.7941656377356624, "noul_auc": 0.7346711447908051, "noul_acc": 0.706, "noul_f1": 0.7699530516431925, "noul_nll": 0.5867154191732407, "noul_ece": 0.0668563278913498, "noul_brier": 0.2000565286770292, "noul_sharp": 0.1619034707546234, "entropy": 0.6146537661552429}
+{"time": "17:35:55", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9814579, "tok_per_s": 806.9924313028692, "rows_per_s": 1.384814012135178, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5691902041435242, "noul_T": 0.8002108628973789, "choice_ce": 0.6705418825149536, "choice_T": 1.0728412258162334, "score_ce": 1.0455740690231323, "score_T": 0.9491235278105069, "noul_auc": 0.7222741333333333, "noul_acc": 0.698, "noul_f1": 0.7739520958083832, "noul_nll": 0.5880829917564988, "noul_ece": 0.04662516665458681, "noul_brier": 0.20118704083871727, "noul_sharp": 0.1676119565963745, "entropy": 0.6107596158981323}, "dropped_long": 0, "peak_gib": 18.236003875732422}

e33/S1-E33-M20P40-n16k-syn/DONE

@@ -0,0 +1 @@
+2026-10-23 11:35:25

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter/README.md

@@ -0,0 +1,206 @@
+---
+base_model: /workspace/jev_kt/models/hf/JEV-9B
+library_name: peft
+tags:
+- base_model:adapter:/workspace/jev_kt/models/hf/JEV-9B
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.21.2

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter/adapter_config.json

@@ -0,0 +1,57 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": "/workspace/jev_kt/models/hf/JEV-9B",
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "out_proj",
+    "v_proj",
+    "in_proj_qkv",
+    "up_proj",
+    "in_proj_z",
+    "down_proj",
+    "gate_proj",
+    "o_proj",
+    "q_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1998c229ad01987725960ed3c5763f1da62efb5d123aeb99b59b9792b7010bce
+size 160486456

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter_vllm/adapter_config.json

@@ -0,0 +1,58 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": null,
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64.0,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "k_proj",
+    "lm_head",
+    "o_proj",
+    "out_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe86a6ba16b8fe092a27422e15e8cbd5f782a3055c61c93e1fb7faf2660766fc
+size 176641832

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.00027095075347460806,
+  -0.0002712052664719522,
+  -0.00215821317397058,
+  -0.0012642444344237447,
+  -0.00037909665843471885,
+  0.0022430443204939365,
+  0.0014937476953491569,
+  -0.0026326365768909454,
+  -0.0004677560064010322,
+  -0.00013563339598476887,
+  0.0005219306913204491,
+  0.0010486760875210166,
+  0.0001400769833708182,
+  -0.0034511531703174114,
+  -0.004273152444511652,
+  -0.005068799946457148,
+  -0.005938724149018526,
+  -0.004366480745375156,
+  -0.004930523689836264,
+  -0.004939631093293428,
+  -0.00384484208188951,
+  -0.00335189257748425,
+  -0.0056142439134418964,
+  -0.003465337445959449
+ ],
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "note": "decision logits = lm_head-LoRA logprobs of verbalizer ids + bias, then per-kind temperature"
+}

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.7536078969778258,
+  "choice": 1.0369940375159572,
+  "score": 0.9641858015101192
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.7536078969778258,
+   "ce_before": 0.5733757615089417,
+   "ce_after": 0.5677251815795898
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0369940375159572,
+   "ce_before": 0.6717521548271179,
+   "ce_after": 0.6714757680892944
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9641858015101192,
+   "ce_before": 1.0550035238265991,
+   "ce_after": 1.0546045303344727
+  }
+ },
+ "source": "models/e33/S1-E33-M20P40-n16k-syn/ckpt-16000"
+}

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:225c17a0e7d3ab4fb832923dcfcc3ccfd1b9835dde3b2f1198f257d7cd21c78e
+size 393464

e33/S1-E33-M20P40-n16k-syn/ckpt-16000/judge_config.json

@@ -0,0 +1,164 @@
+{
+ "base_model_path": "/root/models/Qwen3.5-9B",
+ "hidden_size": 4096,
+ "slots": {
+  "num_slots": 24,
+  "ranges": {
+   "noul": [
+    0,
+    2
+   ],
+   "score": [
+    2,
+    8
+   ],
+   "choice": [
+    8,
+    24
+   ]
+  },
+  "verbalizers": [
+   "false",
+   "true",
+   "0",
+   "1",
+   "2",
+   "3",
+   "4",
+   "5",
+   "A",
+   "B",
+   "C",
+   "D",
+   "E",
+   "F",
+   "G",
+   "H",
+   "I",
+   "J",
+   "K",
+   "L",
+   "M",
+   "N",
+   "O",
+   "P"
+  ],
+  "template_version": "bare-v1"
+ },
+ "verbalizer_ids": [
+  3721,
+  1802,
+  15,
+  16,
+  17,
+  18,
+  19,
+  20,
+  32,
+  33,
+  34,
+  35,
+  36,
+  37,
+  38,
+  39,
+  40,
+  41,
+  42,
+  43,
+  44,
+  45,
+  46,
+  47
+ ],
+ "kinds": [
+  "noul",
+  "choice",
+  "score"
+ ],
+ "model_name": "jev-judge-qwen35-9b",
+ "model_version": "0.8.0",
+ "lm_head_included": true,
+ "weights_mode": "unmerged",
+ "adapter_subfolder": "adapter",
+ "lora": {
+  "r": 16,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "target_modules": [
+   "down_proj",
+   "gate_proj",
+   "in_proj_qkv",
+   "in_proj_z",
+   "k_proj",
+   "o_proj",
+   "out_proj",
+   "q_proj",
+   "up_proj",
+   "v_proj"
+  ]
+ },
+ "trained_stage": "e16",
+ "source_checkpoint": "models/e33/S1-E33-M20P40-n16k-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e33/E33-M20P40-n16k",
+  "data_manifest": {
+   "arm": "E33-M20P40-n16k",
+   "name": null,
+   "train_sha256": "0c998c032f8987adac4577ffcddbc6a65fc3ba6b81ba469c1767c02644b903fb",
+   "n_train": 16000
+  },
+  "init": "released",
+  "objective": "s-log",
+  "sigma": 0.5,
+  "K": 8,
+  "reward": "brier",
+  "clip": 0.2,
+  "passes": 1,
+  "lr": 0.0001,
+  "lr_head": 0.0002,
+  "rows_per_step": 32,
+  "train_seed": 0,
+  "replay": "data/e16/replay/train.jsonl",
+  "replay_frac": 0.05,
+  "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e",
+  "max_len": 3072,
+  "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps",
+  "total_rows": 16000,
+  "lora": {
+   "r": 16,
+   "lora_alpha": 32,
+   "lora_dropout": 0.05
+  },
+  "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3",
+  "event": "ckpt",
+  "tag": "ckpt-16000",
+  "rows_seen": 16000,
+  "replay_seen": 842,
+  "step": 500,
+  "tokens": 9814579,
+  "tok_per_s": 808.7513072822612,
+  "rows_per_s": 1.3878322760350466,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5733757615089417,
+   "noul_T": 0.7536078969778258,
+   "choice_ce": 0.6717521548271179,
+   "choice_T": 1.0369940375159572,
+   "score_ce": 1.0550035238265991,
+   "score_T": 0.9641858015101192,
+   "noul_auc": 0.7241642666666666,
+   "noul_acc": 0.685,
+   "noul_f1": 0.7608200455580866,
+   "noul_nll": 0.590823968693614,
+   "noul_ece": 0.05668055501580238,
+   "noul_brier": 0.2023056829963619,
+   "noul_sharp": 0.15657421946525574,
+   "entropy": 0.6214356422424316
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.236003875732422
+ }
+}

e33/S1-E33-M20P40-n16k-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "08:13:07", "event": "start", "data": "data/e33/E33-M20P40-n16k", "data_manifest": {"arm": "E33-M20P40-n16k", "name": null, "train_sha256": "0c998c032f8987adac4577ffcddbc6a65fc3ba6b81ba469c1767c02644b903fb", "n_train": 16000}, "init": "released", "objective": "s-log", "sigma": 0.5, "K": 8, "reward": "brier", "clip": 0.2, "passes": 1, "lr": 0.0001, "lr_head": 0.0002, "rows_per_step": 32, "train_seed": 0, "replay": "data/e16/replay/train.jsonl", "replay_frac": 0.05, "replay_sha256": "9dd2df19851678750cf79d2fdb47581720930ef2f0c6e18e6fbe23ef2181847e", "max_len": 3072, "schedule": "warmup 0.03 + cosine to x0.05 over 500 steps", "total_rows": 16000, "lora": {"r": 16, "lora_alpha": 32, "lora_dropout": 0.05}, "trainer_sha256": "a0b9d7552588ffb068565686509162a4f122f6f03392e693fa20088493e2a4f3", "device": "cuda", "load_s": 186.75728130340576}
+{"time": "08:16:55", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 834.6921166229316, "loss": 0.6334675438702106, "entropy": 0.6528529115021229, "sharp": 0.11347178183495998}
+{"time": "08:20:44", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 845.8693691529938, "loss": 0.6971936989575624, "entropy": 0.6627442203462124, "sharp": 0.10100687853991985}
+{"time": "08:24:40", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 848.5050223983193, "loss": 0.7298796325922012, "entropy": 0.6808525398373604, "sharp": 0.06191548611968756}
+{"time": "08:28:35", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 850.7955858157326, "loss": 0.7005446255207062, "entropy": 0.639897458255291, "sharp": 0.1383969560265541}
+{"time": "08:32:16", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 850.5229789882324, "loss": 0.6737376227974892, "entropy": 0.6729620024561882, "sharp": 0.08500486146658659}
+{"time": "08:36:13", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 850.9770813230041, "loss": 0.6547166109085083, "entropy": 0.643814604729414, "sharp": 0.12674631737172604}
+{"time": "08:40:00", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 851.2325512337642, "loss": 0.6519396528601646, "entropy": 0.6633624210953712, "sharp": 0.10033598821610212}
+{"time": "08:43:52", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 850.5616989075138, "loss": 0.6236812435090542, "entropy": 0.6104888617992401, "sharp": 0.17594678606837988}
+{"time": "08:47:39", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 849.9966186161708, "loss": 0.6535216812044382, "entropy": 0.6427956651896238, "sharp": 0.13958603981882334}
+{"time": "08:51:31", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 850.7578188501235, "loss": 0.6206827238202095, "entropy": 0.6642791256308556, "sharp": 0.09940087143331766}
+{"time": "08:53:00", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.6004648208618164, "noul_T": 0.7332869469703953, "noul_auc": 0.720477291522506, "noul_acc": 0.708, "noul_f1": 0.7739938080495357, "noul_nll": 0.6005643621385097, "noul_ece": 0.05962640810012819, "noul_brier": 0.20563077495026427, "noul_sharp": 0.14495421946048737, "entropy": 0.6349033117294312}
+{"time": "08:56:37", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 821.824708277265, "loss": 0.6502102166414261, "entropy": 0.6602606698870659, "sharp": 0.10468737687915564}
+{"time": "09:00:21", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 824.2019704334859, "loss": 0.7216617409139872, "entropy": 0.5636089574545622, "sharp": 0.220951359719038}
+{"time": "09:04:07", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 826.0923282324308, "loss": 0.6275180354714394, "entropy": 0.6667738035321236, "sharp": 0.09454494621604681}
+{"time": "09:07:50", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 827.2654550017238, "loss": 0.67619464173913, "entropy": 0.6331211924552917, "sharp": 0.14430847484618425}
+{"time": "09:11:43", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 828.795658712477, "loss": 0.6780807822942734, "entropy": 0.6458774507045746, "sharp": 0.13158006686717272}
+{"time": "09:15:35", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 830.1566643961453, "loss": 0.7391501776874065, "entropy": 0.6528471745550632, "sharp": 0.11345870047807693}
+{"time": "09:19:25", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 831.4080472090963, "loss": 0.6620825305581093, "entropy": 0.6246006488800049, "sharp": 0.16433628741651773}
+{"time": "09:23:19", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 832.2049488245029, "loss": 0.6418737918138504, "entropy": 0.6729150265455246, "sharp": 0.07942454144358635}
+{"time": "09:27:07", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 833.2000802393027, "loss": 0.6191496253013611, "entropy": 0.6573431305587292, "sharp": 0.11025342717766762}
+{"time": "09:30:59", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 834.2327115681361, "loss": 0.6632450968027115, "entropy": 0.6655780747532845, "sharp": 0.08871013019233942}
+{"time": "09:32:28", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.598476767539978, "noul_T": 0.7084120180784974, "noul_auc": 0.7335741168925418, "noul_acc": 0.692, "noul_f1": 0.7441860465116279, "noul_nll": 0.5992745052427053, "noul_ece": 0.0875435711145401, "noul_brier": 0.2056747248190986, "noul_sharp": 0.1401194930076599, "entropy": 0.6354633569717407}
+{"time": "09:36:20", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 820.6502771891819, "loss": 0.6477267909795046, "entropy": 0.66724544018507, "sharp": 0.09406618867069483}
+{"time": "09:40:07", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 822.0828850419384, "loss": 0.6350954649969935, "entropy": 0.645612420514226, "sharp": 0.13076961133629084}
+{"time": "09:44:04", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 823.0521579520484, "loss": 0.6120151542127132, "entropy": 0.634583511389792, "sharp": 0.14943482168018818}
+{"time": "09:48:00", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 824.0820127062603, "loss": 0.6576330363750458, "entropy": 0.6363804861903191, "sharp": 0.1506327036768198}
+{"time": "09:51:54", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 825.1990614839201, "loss": 0.6741576343774796, "entropy": 0.6751580238342285, "sharp": 0.07021459378302097}
+{"time": "09:55:37", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 826.3026104339746, "loss": 0.6422720225527883, "entropy": 0.6670975536108017, "sharp": 0.09885596670210361}
+{"time": "09:59:28", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 827.375040384621, "loss": 0.6690735965967178, "entropy": 0.6420081779360771, "sharp": 0.1370291942730546}
+{"time": "10:03:17", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 828.4143308609, "loss": 0.7405221462249756, "entropy": 0.6498714536428452, "sharp": 0.10969929583370686}
+{"time": "10:07:08", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 829.3279071091144, "loss": 0.6922129094600677, "entropy": 0.6504716016352177, "sharp": 0.12393020559102297}
+{"time": "10:10:56", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 830.0992317733071, "loss": 0.6168073639273643, "entropy": 0.6522897556424141, "sharp": 0.11487219575792551}
+{"time": "10:12:25", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.5929418206214905, "noul_T": 0.6742968264197762, "noul_auc": 0.7332703245514842, "noul_acc": 0.71, "noul_f1": 0.7672552166934189, "noul_nll": 0.5955979643315077, "noul_ece": 0.084287614107132, "noul_brier": 0.20386416261754284, "noul_sharp": 0.13998760282993317, "entropy": 0.6352567672729492}
+{"time": "10:16:13", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 820.8849739366888, "loss": 0.6018178090453148, "entropy": 0.6624333262443542, "sharp": 0.10367251187562943}
+{"time": "10:20:06", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 822.1702387415982, "loss": 0.6465166732668877, "entropy": 0.6494022682309151, "sharp": 0.1233661575242877}
+{"time": "10:23:53", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 823.4054480792602, "loss": 0.5945486165583134, "entropy": 0.6228913813829422, "sharp": 0.15881608240306377}
+{"time": "10:27:40", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 824.4701371797404, "loss": 0.6590227223932743, "entropy": 0.6720576547086239, "sharp": 0.08543357811868191}
+{"time": "10:31:31", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 825.54706766103, "loss": 0.6911755874752998, "entropy": 0.6557943299412727, "sharp": 0.11337604280561209}
+{"time": "10:35:16", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 826.4094436168443, "loss": 0.672113474458456, "entropy": 0.6336255744099617, "sharp": 0.13487344607710838}
+{"time": "10:39:03", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 827.1306496519461, "loss": 0.6700553707778454, "entropy": 0.6551447063684464, "sharp": 0.11386012751609087}
+{"time": "10:42:56", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 827.8117141715628, "loss": 0.6640688851475716, "entropy": 0.6413109470158815, "sharp": 0.13535328581929207}
+{"time": "10:46:47", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 828.5516928774163, "loss": 0.6268397979438305, "entropy": 0.632170895114541, "sharp": 0.14578355848789215}
+{"time": "10:50:38", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 829.2252069550974, "loss": 0.5890467502176762, "entropy": 0.644784465432167, "sharp": 0.12795494962483644}
+{"time": "10:52:07", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5862334966659546, "noul_T": 0.7606483978429152, "noul_auc": 0.7381647567129669, "noul_acc": 0.7, "noul_f1": 0.7706422018348624, "noul_nll": 0.5874892692416906, "noul_ece": 0.0580228523015976, "noul_brier": 0.20039999223207658, "noul_sharp": 0.1556634157896042, "entropy": 0.6206153035163879}
+{"time": "10:55:54", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 822.4939761646747, "loss": 0.6346560120582581, "entropy": 0.6383396536111832, "sharp": 0.1300529083237052}
+{"time": "10:59:44", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 823.1133792883301, "loss": 0.6326565071940422, "entropy": 0.6483790576457977, "sharp": 0.10434569697827101}
+{"time": "11:03:35", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 823.9389462869989, "loss": 0.6603165343403816, "entropy": 0.6323894076049328, "sharp": 0.141101255081594}
+{"time": "11:07:27", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 824.6522983284796, "loss": 0.6576901823282242, "entropy": 0.6554438918828964, "sharp": 0.10777606535702944}
+{"time": "11:11:16", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 825.4080300941106, "loss": 0.6637649387121201, "entropy": 0.6494643464684486, "sharp": 0.12167353462427855}
+{"time": "11:15:02", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 826.0346436111113, "loss": 0.6434320770204067, "entropy": 0.6648931764066219, "sharp": 0.09884033910930157}
+{"time": "11:19:00", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 826.328165731913, "loss": 0.6828350089490414, "entropy": 0.6499973721802235, "sharp": 0.12077272217720747}
+{"time": "11:23:00", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 826.7948135441493, "loss": 0.6283932998776436, "entropy": 0.641958087682724, "sharp": 0.13573688734322786}
+{"time": "11:26:51", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 827.1983496490172, "loss": 0.6231705062091351, "entropy": 0.6532550044357777, "sharp": 0.11139336042106152}
+{"time": "11:30:45", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 827.6767047919681, "loss": 0.6067280508577824, "entropy": 0.6390356905758381, "sharp": 0.14126633014529943}
+{"time": "11:32:14", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5895894765853882, "noul_T": 0.7579464023116482, "noul_auc": 0.737270257042075, "noul_acc": 0.696, "noul_f1": 0.7610062893081762, "noul_nll": 0.5895842367261648, "noul_ece": 0.05830827236175538, "noul_brier": 0.20119680086413858, "noul_sharp": 0.15372470021247864, "entropy": 0.6235913038253784}
+{"time": "11:35:25", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9814579, "tok_per_s": 808.7513072822612, "rows_per_s": 1.3878322760350466, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5733757615089417, "noul_T": 0.7536078969778258, "choice_ce": 0.6717521548271179, "choice_T": 1.0369940375159572, "score_ce": 1.0550035238265991, "score_T": 0.9641858015101192, "noul_auc": 0.7241642666666666, "noul_acc": 0.685, "noul_f1": 0.7608200455580866, "noul_nll": 0.590823968693614, "noul_ece": 0.05668055501580238, "noul_brier": 0.2023056829963619, "noul_sharp": 0.15657421946525574, "entropy": 0.6214356422424316}, "dropped_long": 0, "peak_gib": 18.236003875732422}

e33/S1-E33-M40P40-n16k-a100-syn/DONE

@@ -0,0 +1 @@
+2026-10-24 06:50:10

e33/S1-E33-M40P40-n16k-a100-syn/ckpt-16000/adapter/README.md

@@ -0,0 +1,206 @@
+---
+base_model: /workspace/jev_kt/models/hf/JEV-9B
+library_name: peft
+tags:
+- base_model:adapter:/workspace/jev_kt/models/hf/JEV-9B
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.21.2

e33/S1-E33-M40P40-n16k-a100-syn/ckpt-16000/adapter/adapter_config.json

@@ -0,0 +1,57 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": "/workspace/jev_kt/models/hf/JEV-9B",
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "o_proj",
+    "in_proj_z",
+    "k_proj",
+    "up_proj",
+    "v_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "out_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M40P40-n16k-a100-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c39fecddf4eb407c1d1f604409131c1052549d8d610d0b6ee185e15b8a1e3980
+size 160486456

e33/S1-E33-M40P40-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_config.json

@@ -0,0 +1,58 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen3_5ForCausalLM",
+    "parent_library": "transformers.models.qwen3_5.modeling_qwen3_5"
+  },
+  "base_model_name_or_path": null,
+  "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,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64.0,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.0",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "k_proj",
+    "lm_head",
+    "o_proj",
+    "out_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": null,
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

e33/S1-E33-M40P40-n16k-a100-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0779ab950735755fe718df27c1dbad733a41375c4fac3bdbb3157d77c5df16f8
+size 176641832