Pod D e28 checkpoints (E27 27B / E28 diversity / E29 / E30 arms)

e28/S1-E28-C0-syn/DONE

@@ -0,0 +1 @@
+2026-10-20 06:33:13

e28/S1-E28-C0-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

e28/S1-E28-C0-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": [
+    "up_proj",
+    "o_proj",
+    "v_proj",
+    "in_proj_qkv",
+    "k_proj",
+    "gate_proj",
+    "out_proj",
+    "q_proj",
+    "in_proj_z",
+    "down_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
+}

e28/S1-E28-C0-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea15bc4fc52ad521385fee6a115227c0b50e9a7164e2578a47cc48e63362d083
+size 160486456

e28/S1-E28-C0-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
+}

e28/S1-E28-C0-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4805bbcf619946a680b929f9c21e8dbd0d9ba163167dffb7cc13dd8182303fce
+size 176641832

e28/S1-E28-C0-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.0004019341722596437,
+  -0.00040218859794549644,
+  -0.0023606743197888136,
+  -0.0017429249128326774,
+  -0.0005036078509874642,
+  0.0018798642558977008,
+  0.0014552718494087458,
+  -0.002158828778192401,
+  8.943251305026934e-05,
+  -7.114288746379316e-05,
+  -5.446747763926396e-07,
+  0.0007199719548225403,
+  -0.0017304084030911326,
+  -0.003423966933041811,
+  -0.004143770784139633,
+  -0.005592531058937311,
+  -0.005568793509155512,
+  -0.00443001464009285,
+  -0.0049525978974998,
+  -0.005071173422038555,
+  -0.003785105422139168,
+  -0.003439312567934394,
+  -0.005798427853733301,
+  -0.0035148358438163996
+ ],
+ "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"
+}

e28/S1-E28-C0-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.8210398824005378,
+  "choice": 1.0076231342175517,
+  "score": 0.9955852301156168
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.8210398824005378,
+   "ce_before": 0.5733858942985535,
+   "ce_after": 0.5705401301383972
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0076231342175517,
+   "ce_before": 0.6707404851913452,
+   "ce_after": 0.6707300543785095
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9955852301156168,
+   "ce_before": 1.055969476699829,
+   "ce_after": 1.0559628009796143
+  }
+ },
+ "source": "models/e28/S1-E28-C0-syn/ckpt-16000"
+}

e28/S1-E28-C0-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c7b7a054dfd9924cb58a4007d7caea6920434919ef35e46b9e5daf7c0fefb66
+size 393464

e28/S1-E28-C0-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/e28/S1-E28-C0-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e19/E19-b-n16k",
+  "data_manifest": {
+   "arm": "E19-b-n16k",
+   "name": null,
+   "train_sha256": "9e31412d9aac0064859cdf9aae1636242d1a9dbfc3ab398a1f5ec10da1e7a50e",
+   "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": 9808867,
+  "tok_per_s": 3237.0095425583345,
+  "rows_per_s": 5.558003251258509,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5733858942985535,
+   "noul_T": 0.8210398824005378,
+   "choice_ce": 0.6707404851913452,
+   "choice_T": 1.0076231342175517,
+   "score_ce": 1.055969476699829,
+   "score_T": 0.9955852301156168,
+   "noul_auc": 0.7198933333333333,
+   "noul_acc": 0.699,
+   "noul_f1": 0.7794871794871795,
+   "noul_nll": 0.5898938082531094,
+   "noul_ece": 0.052008005946874616,
+   "noul_brier": 0.20208435675807834,
+   "noul_sharp": 0.1651003509759903,
+   "entropy": 0.6122925281524658
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.282634735107422
+ }
+}

e28/S1-E28-C0-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "05:42:42", "event": "start", "data": "data/e19/E19-b-n16k", "data_manifest": {"arm": "E19-b-n16k", "name": null, "train_sha256": "9e31412d9aac0064859cdf9aae1636242d1a9dbfc3ab398a1f5ec10da1e7a50e", "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": 4.794634580612183}
+{"time": "05:44:34", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 1742.0688985934723, "loss": 0.6418524533510208, "entropy": 0.6674938574433327, "sharp": 0.09449283499270678}
+{"time": "05:45:28", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 2276.391236296654, "loss": 0.6543252449482679, "entropy": 0.6451239045709372, "sharp": 0.11809633672237396}
+{"time": "05:46:27", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 2592.524926228113, "loss": 0.7097902707755566, "entropy": 0.6600528657436371, "sharp": 0.11035785730928183}
+{"time": "05:47:22", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 2768.586496905704, "loss": 0.6805907804518938, "entropy": 0.6736207865178585, "sharp": 0.08694715239107609}
+{"time": "05:48:18", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 2889.323608493551, "loss": 0.6551038809120655, "entropy": 0.6535464469343424, "sharp": 0.10296877473592758}
+{"time": "05:49:14", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 2977.4619186367863, "loss": 0.6448147781193256, "entropy": 0.6451537758111954, "sharp": 0.12875191681087017}
+{"time": "05:50:12", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 3036.6014164640724, "loss": 0.647441454231739, "entropy": 0.683087445795536, "sharp": 0.057668830268085}
+{"time": "05:51:08", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 3084.627950729468, "loss": 0.7112917471677065, "entropy": 0.637435607612133, "sharp": 0.1373877078294754}
+{"time": "05:52:04", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 3124.1413080678926, "loss": 0.6655900441110134, "entropy": 0.6652580164372921, "sharp": 0.10431478638201952}
+{"time": "05:53:00", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 3152.110435086271, "loss": 0.6226409263908863, "entropy": 0.6681161969900131, "sharp": 0.09350143652409315}
+{"time": "05:53:25", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.6027932167053223, "noul_T": 0.8047402751192091, "noul_auc": 0.7031611280822265, "noul_acc": 0.688, "noul_f1": 0.75625, "noul_nll": 0.6076746896356344, "noul_ece": 0.06157691490650178, "noul_brier": 0.2100360241730981, "noul_sharp": 0.14764730632305145, "entropy": 0.6277173757553101}
+{"time": "05:54:23", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 3068.6206931182887, "loss": 0.684677317738533, "entropy": 0.6516259256750345, "sharp": 0.11634835135191679}
+{"time": "05:55:19", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 3089.830183059434, "loss": 0.6259176284074783, "entropy": 0.6395060457289219, "sharp": 0.14079864881932735}
+{"time": "05:56:16", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 3116.7881936735016, "loss": 0.6442239172756672, "entropy": 0.6784866154193878, "sharp": 0.06614390201866627}
+{"time": "05:57:14", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 3139.759900493683, "loss": 0.6844842508435249, "entropy": 0.64948820322752, "sharp": 0.13435495272278786}
+{"time": "05:58:09", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 3160.5145456206246, "loss": 0.6813701651990414, "entropy": 0.668898917734623, "sharp": 0.08269481174647808}
+{"time": "05:59:06", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 3179.978036598646, "loss": 0.7068287283182144, "entropy": 0.6260516047477722, "sharp": 0.15484808944165707}
+{"time": "06:00:04", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 3196.797926526836, "loss": 0.6369537115097046, "entropy": 0.6493389010429382, "sharp": 0.11900342628359795}
+{"time": "06:01:00", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 3211.968891141825, "loss": 0.6736315377056599, "entropy": 0.6538582667708397, "sharp": 0.11945641413331032}
+{"time": "06:01:56", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 3223.8353887126264, "loss": 0.6503072865307331, "entropy": 0.6558262407779694, "sharp": 0.10804163292050362}
+{"time": "06:02:53", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 3235.046518412555, "loss": 0.6812480166554451, "entropy": 0.6733076274394989, "sharp": 0.07814495917409658}
+{"time": "06:03:13", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.625215470790863, "noul_T": 0.6322796617575723, "noul_auc": 0.7433798585677879, "noul_acc": 0.66, "noul_f1": 0.6875, "noul_nll": 0.6210226207077503, "noul_ece": 0.14127235460281373, "noul_brier": 0.21537573031776094, "noul_sharp": 0.11072076112031937, "entropy": 0.6588683724403381}
+{"time": "06:04:09", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 3195.1610605836995, "loss": 0.6731307283043861, "entropy": 0.6324892044067383, "sharp": 0.14314057677984238}
+{"time": "06:05:04", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 3207.7340288322866, "loss": 0.7052969560027122, "entropy": 0.6550907492637634, "sharp": 0.11679293122142553}
+{"time": "06:06:01", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 3218.423092980036, "loss": 0.6667746938765049, "entropy": 0.6606782749295235, "sharp": 0.10835204273462296}
+{"time": "06:06:58", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 3228.7448099209337, "loss": 0.6714807525277138, "entropy": 0.6531186066567898, "sharp": 0.11655727680772543}
+{"time": "06:07:54", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 3238.0592523581136, "loss": 0.6987750642001629, "entropy": 0.6320639178156853, "sharp": 0.15557408891618252}
+{"time": "06:08:50", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 3245.5095129602087, "loss": 0.653316430747509, "entropy": 0.6700648739933968, "sharp": 0.0919972499832511}
+{"time": "06:09:49", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 3255.199106196958, "loss": 0.640799886547029, "entropy": 0.6467154920101166, "sharp": 0.13283592090010643}
+{"time": "06:10:46", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 3262.485544814, "loss": 0.6944353505969048, "entropy": 0.6748707816004753, "sharp": 0.08099196664988995}
+{"time": "06:11:43", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 3270.3215016883805, "loss": 0.6646610796451569, "entropy": 0.6652002595365047, "sharp": 0.0954577811062336}
+{"time": "06:12:41", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 3276.736460884371, "loss": 0.6731904521584511, "entropy": 0.6602216698229313, "sharp": 0.11017826851457357}
+{"time": "06:13:01", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.597034752368927, "noul_T": 0.6421206855787178, "noul_auc": 0.7323589475283117, "noul_acc": 0.718, "noul_f1": 0.7860394537177542, "noul_nll": 0.5973743291497231, "noul_ece": 0.08663646721839906, "noul_brier": 0.2042407758885801, "noul_sharp": 0.1332443803548813, "entropy": 0.6418294906616211}
+{"time": "06:13:56", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 3247.7827177374384, "loss": 0.6042209938168526, "entropy": 0.6657007820904255, "sharp": 0.10047882981598377}
+{"time": "06:14:53", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 3254.074569995877, "loss": 0.672166720032692, "entropy": 0.6612759456038475, "sharp": 0.09925397951155901}
+{"time": "06:15:51", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 3260.7679436398503, "loss": 0.6972816735506058, "entropy": 0.643512986600399, "sharp": 0.12919290456920862}
+{"time": "06:16:45", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 3266.4804837555394, "loss": 0.6724905520677567, "entropy": 0.6493063047528267, "sharp": 0.11664058640599251}
+{"time": "06:17:42", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 3272.1660910371143, "loss": 0.6453768238425255, "entropy": 0.6563559770584106, "sharp": 0.11054976005107164}
+{"time": "06:18:38", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 3277.0396115841886, "loss": 0.6560497954487801, "entropy": 0.635651208460331, "sharp": 0.14939494617283344}
+{"time": "06:19:35", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 3282.9871352353107, "loss": 0.686661321669817, "entropy": 0.6401583924889565, "sharp": 0.13821567595005035}
+{"time": "06:20:32", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 3287.3508514962205, "loss": 0.6733216270804405, "entropy": 0.6461369395256042, "sharp": 0.12509455811232328}
+{"time": "06:21:29", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 3293.099189016152, "loss": 0.6431321948766708, "entropy": 0.6418887823820114, "sharp": 0.1216413863003254}
+{"time": "06:22:23", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 3296.8323002762377, "loss": 0.6882326304912567, "entropy": 0.6303502470254898, "sharp": 0.14134303573518991}
+{"time": "06:22:43", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5883752703666687, "noul_T": 0.8063948425248696, "noul_auc": 0.7352956068252012, "noul_acc": 0.722, "noul_f1": 0.7851622874806801, "noul_nll": 0.5876846368610859, "noul_ece": 0.05814056313037875, "noul_brier": 0.20021823288795024, "noul_sharp": 0.1605500727891922, "entropy": 0.6165220141410828}
+{"time": "06:23:40", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 3273.217062149899, "loss": 0.6581775918602943, "entropy": 0.6659778021275997, "sharp": 0.09172998555004597}
+{"time": "06:24:35", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 3278.0760739602315, "loss": 0.6624073907732964, "entropy": 0.6563186943531036, "sharp": 0.10841984767466784}
+{"time": "06:25:33", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 3282.822632861716, "loss": 0.668071087449789, "entropy": 0.646204873919487, "sharp": 0.12354523129761219}
+{"time": "06:26:29", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 3287.3099151168462, "loss": 0.6683605443686247, "entropy": 0.6525228191167116, "sharp": 0.11701337993144989}
+{"time": "06:27:26", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 3291.6826129417955, "loss": 0.6479352544993162, "entropy": 0.6605901643633842, "sharp": 0.09802819322794676}
+{"time": "06:28:22", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 3295.468233519694, "loss": 0.6506122052669525, "entropy": 0.6533954069018364, "sharp": 0.11456773895770311}
+{"time": "06:29:20", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 3299.817314706129, "loss": 0.7218317352235317, "entropy": 0.6592727117240429, "sharp": 0.10681359004229307}
+{"time": "06:30:16", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 3303.496486455061, "loss": 0.6903524547815323, "entropy": 0.6571662090718746, "sharp": 0.10609687305986881}
+{"time": "06:31:10", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 3306.3984164704566, "loss": 0.5719492360949516, "entropy": 0.6424555853009224, "sharp": 0.13239238318055868}
+{"time": "06:32:06", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 3309.895620939873, "loss": 0.6167443245649338, "entropy": 0.6454180404543877, "sharp": 0.12842741049826145}
+{"time": "06:32:26", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5884483456611633, "noul_T": 0.798413540070553, "noul_auc": 0.7332534471992033, "noul_acc": 0.72, "noul_f1": 0.7859327217125383, "noul_nll": 0.5891514290720224, "noul_ece": 0.061191798210144054, "noul_brier": 0.200934913549001, "noul_sharp": 0.15952661633491516, "entropy": 0.6176946759223938}
+{"time": "06:33:13", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9808867, "tok_per_s": 3237.0095425583345, "rows_per_s": 5.558003251258509, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5733858942985535, "noul_T": 0.8210398824005378, "choice_ce": 0.6707404851913452, "choice_T": 1.0076231342175517, "score_ce": 1.055969476699829, "score_T": 0.9955852301156168, "noul_auc": 0.7198933333333333, "noul_acc": 0.699, "noul_f1": 0.7794871794871795, "noul_nll": 0.5898938082531094, "noul_ece": 0.052008005946874616, "noul_brier": 0.20208435675807834, "noul_sharp": 0.1651003509759903, "entropy": 0.6122925281524658}, "dropped_long": 0, "peak_gib": 18.282634735107422}

e28/S1-E28-E18A050-rt-syn/DONE

@@ -0,0 +1 @@
+2026-10-21 06:54:56

e28/S1-E28-E18A050-rt-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

e28/S1-E28-E18A050-rt-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": [
+    "o_proj",
+    "k_proj",
+    "in_proj_z",
+    "out_proj",
+    "q_proj",
+    "v_proj",
+    "up_proj",
+    "down_proj",
+    "in_proj_qkv",
+    "gate_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
+}

e28/S1-E28-E18A050-rt-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d32e8cade04cb62e764ce8960d6d4340906a030b264367825ba9b79963a9929
+size 160486456

e28/S1-E28-E18A050-rt-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
+}

e28/S1-E28-E18A050-rt-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb2307e1988ab4b5c8524539ee27ea8a645102f2a10ab1c486095071391da9a6
+size 176641832

e28/S1-E28-E18A050-rt-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.00039831086178310215,
+  -0.00039857125375419855,
+  -0.0030545094050467014,
+  -0.0019210963509976864,
+  -0.0006272284663282335,
+  0.0020584675949066877,
+  0.0012770768953487277,
+  -0.003080836497247219,
+  -0.00024181828484870493,
+  -0.0002333867159904912,
+  0.00029039953369647264,
+  0.0007457923493348062,
+  0.00015367088781204075,
+  -0.0038240239955484867,
+  -0.00451631611213088,
+  -0.005286028143018484,
+  -0.006027279421687126,
+  -0.004458629060536623,
+  -0.004973826929926872,
+  -0.005131916608661413,
+  -0.0040062107145786285,
+  -0.003536548698320985,
+  -0.005771616939455271,
+  -0.003572249785065651
+ ],
+ "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"
+}

e28/S1-E28-E18A050-rt-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.8495371650607102,
+  "choice": 1.0560892190772713,
+  "score": 0.960112119036335
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.8495371650607102,
+   "ce_before": 0.5595800280570984,
+   "ce_after": 0.5574237108230591
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.0560892190772713,
+   "ce_before": 0.6842275261878967,
+   "ce_after": 0.683573842048645
+  },
+  "score": {
+   "n": 53,
+   "T": 0.960112119036335,
+   "ce_before": 1.068756341934204,
+   "ce_after": 1.0682735443115234
+  }
+ },
+ "source": "models/e28/S1-E28-E18A050-rt-syn/ckpt-16000"
+}

e28/S1-E28-E18A050-rt-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04f1e3aebba797e65e6f251cf88d134d9cb90ba13166ecaa4ef8d3113bfbdf1d
+size 393464

e28/S1-E28-E18A050-rt-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/e28/S1-E28-E18A050-rt-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e28/E28-E18A050-rt",
+  "data_manifest": {
+   "arm": "G-d512-n16k",
+   "name": null,
+   "train_sha256": "ef94315960694ce58bdc6d61b8665b7bf46f4e2996b9e82503e1d17bb87a14b7",
+   "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": 9113372,
+  "tok_per_s": 3214.371023602123,
+  "rows_per_s": 5.940329960631729,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5595800280570984,
+   "noul_T": 0.8495371650607102,
+   "choice_ce": 0.6842275261878967,
+   "choice_T": 1.0560892190772713,
+   "score_ce": 1.068756341934204,
+   "score_T": 0.960112119036335,
+   "noul_auc": 0.7447338666666667,
+   "noul_acc": 0.695,
+   "noul_f1": 0.7677075399847677,
+   "noul_nll": 0.5717920997068286,
+   "noul_ece": 0.04427964748442174,
+   "noul_brier": 0.19495265463736913,
+   "noul_sharp": 0.1802748143672943,
+   "entropy": 0.5956453680992126
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.284047603607178
+ }
+}

e28/S1-E28-E18A050-rt-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "06:07:41", "event": "start", "data": "data/e28/E28-E18A050-rt", "data_manifest": {"arm": "G-d512-n16k", "name": null, "train_sha256": "ef94315960694ce58bdc6d61b8665b7bf46f4e2996b9e82503e1d17bb87a14b7", "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": 4.799657821655273}
+{"time": "06:08:35", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 3335.8686490784153, "loss": 0.6002982165664434, "entropy": 0.5942575950175524, "sharp": 0.19537658989429474}
+{"time": "06:09:29", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 3347.89659647421, "loss": 0.6112224832177162, "entropy": 0.59024703130126, "sharp": 0.2019644845277071}
+{"time": "06:10:23", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 3353.529258976685, "loss": 0.6076782438904047, "entropy": 0.529285728931427, "sharp": 0.24768979381769896}
+{"time": "06:11:16", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 3359.1094743060803, "loss": 0.5808004774153233, "entropy": 0.6097748354077339, "sharp": 0.17353096511214972}
+{"time": "06:12:11", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 3369.3068844360805, "loss": 0.5585350655019283, "entropy": 0.5976961776614189, "sharp": 0.1846163496375084}
+{"time": "06:13:04", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 3375.8048401322844, "loss": 0.5941595509648323, "entropy": 0.6318124160170555, "sharp": 0.154153554700315}
+{"time": "06:13:58", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 3375.8535966310665, "loss": 0.5850520320236683, "entropy": 0.5771617069840431, "sharp": 0.20421438571065664}
+{"time": "06:14:52", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 3377.0916069751543, "loss": 0.6283475775271654, "entropy": 0.601615758612752, "sharp": 0.18012536317110062}
+{"time": "06:15:47", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 3380.7041022056424, "loss": 0.6425563246011734, "entropy": 0.594768863171339, "sharp": 0.17997646797448397}
+{"time": "06:16:41", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 3380.585620079485, "loss": 0.5740486830472946, "entropy": 0.6155685745179653, "sharp": 0.16669677570462227}
+{"time": "06:17:01", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.5806242823600769, "noul_T": 0.9017193599054513, "noul_auc": 0.7404600766231794, "noul_acc": 0.708, "noul_f1": 0.7652733118971061, "noul_nll": 0.5830994355976582, "noul_ece": 0.06945477390289306, "noul_brier": 0.1993663816604694, "noul_sharp": 0.1820717602968216, "entropy": 0.5966274738311768}
+{"time": "06:17:54", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 3274.935834648614, "loss": 0.6264102570712566, "entropy": 0.6102959588170052, "sharp": 0.16732802893966436}
+{"time": "06:18:50", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 3287.0189864821386, "loss": 0.5803467445075512, "entropy": 0.5993821527808905, "sharp": 0.17582901380956173}
+{"time": "06:19:47", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 3297.0851454234808, "loss": 0.5702192187309265, "entropy": 0.5612410046160221, "sharp": 0.21354103833436966}
+{"time": "06:20:41", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 3302.716488321634, "loss": 0.6200642455369234, "entropy": 0.6473408713936806, "sharp": 0.12580503337085247}
+{"time": "06:21:36", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 3308.2881403457145, "loss": 0.6428153924643993, "entropy": 0.6326981969177723, "sharp": 0.14672968070954084}
+{"time": "06:22:28", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 3312.8328430870833, "loss": 0.6297238897532225, "entropy": 0.5767229981720448, "sharp": 0.19916999712586403}
+{"time": "06:23:23", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 3318.1738265316453, "loss": 0.6130420938134193, "entropy": 0.6221097409725189, "sharp": 0.15260346792638302}
+{"time": "06:24:17", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 3322.7193927372523, "loss": 0.5934165082871914, "entropy": 0.6152561269700527, "sharp": 0.16756878979504108}
+{"time": "06:25:11", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 3324.589217224558, "loss": 0.6296769827604294, "entropy": 0.6074256487190723, "sharp": 0.16790608409792185}
+{"time": "06:26:05", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 3328.6099107837185, "loss": 0.5907804258167744, "entropy": 0.6081702299416065, "sharp": 0.18031224142760038}
+{"time": "06:26:25", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.5784547328948975, "noul_T": 0.8328414215411776, "noul_auc": 0.7448313108639516, "noul_acc": 0.704, "noul_f1": 0.7620578778135049, "noul_nll": 0.5811050497889518, "noul_ece": 0.06479078799486161, "noul_brier": 0.19810341077413438, "noul_sharp": 0.17559659481048584, "entropy": 0.6053183078765869}
+{"time": "06:27:20", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 3274.7058554089595, "loss": 0.586168397217989, "entropy": 0.5733136478811502, "sharp": 0.20715579017996788}
+{"time": "06:28:13", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 3278.7195425829236, "loss": 0.6014519557356834, "entropy": 0.593541793525219, "sharp": 0.18261277489364147}
+{"time": "06:29:06", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 3282.2559446384985, "loss": 0.5910805948078632, "entropy": 0.5957015547901392, "sharp": 0.19531946256756783}
+{"time": "06:29:59", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 3285.9311897875646, "loss": 0.5925949309021235, "entropy": 0.5877530910074711, "sharp": 0.19521856494247913}
+{"time": "06:30:57", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 3288.701294556637, "loss": 0.6153495758771896, "entropy": 0.6061336323618889, "sharp": 0.1711768750101328}
+{"time": "06:31:48", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 3291.160656994244, "loss": 0.5898217596113682, "entropy": 0.6045204438269138, "sharp": 0.16672593727707863}
+{"time": "06:32:41", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 3294.217945249048, "loss": 0.6414357991889119, "entropy": 0.6187189966440201, "sharp": 0.15950102172791958}
+{"time": "06:33:34", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 3298.173177428458, "loss": 0.6342934807762504, "entropy": 0.6112399101257324, "sharp": 0.17046577669680119}
+{"time": "06:34:29", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 3301.4133787114624, "loss": 0.617023091763258, "entropy": 0.5880597829818726, "sharp": 0.18076064437627792}
+{"time": "06:35:24", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 3304.863377969232, "loss": 0.5869416482746601, "entropy": 0.5788372401148081, "sharp": 0.1744002364575863}
+{"time": "06:35:44", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.5695884227752686, "noul_T": 0.8970043922903317, "noul_auc": 0.7486624698317328, "noul_acc": 0.706, "noul_f1": 0.7692307692307693, "noul_nll": 0.5728995004072786, "noul_ece": 0.03249754613637924, "noul_brier": 0.19544740658234394, "noul_sharp": 0.18595445156097412, "entropy": 0.586853563785553}
+{"time": "06:36:36", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 3268.4624207100774, "loss": 0.5773440040647984, "entropy": 0.6028434857726097, "sharp": 0.16967926733195782}
+{"time": "06:37:27", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 3271.3165014306455, "loss": 0.6374371545389295, "entropy": 0.6155262598767877, "sharp": 0.16538868471980095}
+{"time": "06:38:21", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 3275.03692884936, "loss": 0.5824361257255077, "entropy": 0.5844896733760834, "sharp": 0.18815483525395393}
+{"time": "06:39:15", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 3278.093736059273, "loss": 0.6290938165038824, "entropy": 0.6002766862511635, "sharp": 0.17840714938938618}
+{"time": "06:40:06", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 3281.1930558265444, "loss": 0.6121017262339592, "entropy": 0.61546690762043, "sharp": 0.14994669798761606}
+{"time": "06:40:59", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 3284.7178698010935, "loss": 0.6235643997788429, "entropy": 0.598175898194313, "sharp": 0.18592064268887043}
+{"time": "06:41:52", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 3288.22198246336, "loss": 0.6016076346859336, "entropy": 0.5771215986460447, "sharp": 0.20257375203073025}
+{"time": "06:42:47", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 3290.5559815850797, "loss": 0.6171148121356964, "entropy": 0.6074099205434322, "sharp": 0.1688613723963499}
+{"time": "06:43:43", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 3294.047448880535, "loss": 0.6298759374767542, "entropy": 0.6090978160500526, "sharp": 0.16564554441720247}
+{"time": "06:44:35", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 3295.362348553038, "loss": 0.6032796315848827, "entropy": 0.607449185103178, "sharp": 0.16136501543223858}
+{"time": "06:44:55", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5710271000862122, "noul_T": 0.8856563235718324, "noul_auc": 0.7491350356955999, "noul_acc": 0.712, "noul_f1": 0.7770897832817337, "noul_nll": 0.5730964285135269, "noul_ece": 0.03887997013330461, "noul_brier": 0.19541467243427132, "noul_sharp": 0.18427079916000366, "entropy": 0.590028703212738}
+{"time": "06:45:48", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 3268.2654171836684, "loss": 0.6083678230643272, "entropy": 0.602668821811676, "sharp": 0.17243992909789085}
+{"time": "06:46:40", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 3269.8583601611954, "loss": 0.6399169377982616, "entropy": 0.6169490301981568, "sharp": 0.15903472062200308}
+{"time": "06:47:35", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 3272.142937401324, "loss": 0.612063997425139, "entropy": 0.5993046313524246, "sharp": 0.17189513705670834}
+{"time": "06:48:27", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 3274.5345470432685, "loss": 0.6128622889518738, "entropy": 0.6048808097839355, "sharp": 0.16242566797882318}
+{"time": "06:49:23", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 3276.7428045035026, "loss": 0.5755896419286728, "entropy": 0.6024290472269058, "sharp": 0.1778944954276085}
+{"time": "06:50:14", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 3278.459756061006, "loss": 0.5963110364973545, "entropy": 0.5952213108539581, "sharp": 0.18893328122794628}
+{"time": "06:51:06", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 3280.259019363613, "loss": 0.6257474981248379, "entropy": 0.6058809459209442, "sharp": 0.18065770715475082}
+{"time": "06:52:02", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 3282.658107966944, "loss": 0.6027997732162476, "entropy": 0.5877076126635075, "sharp": 0.1966692004352808}
+{"time": "06:52:58", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 3284.523572077551, "loss": 0.5845214314758778, "entropy": 0.5873317420482635, "sharp": 0.1952001266181469}
+{"time": "06:53:53", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 3286.6416338650647, "loss": 0.5627817288041115, "entropy": 0.5550961457192898, "sharp": 0.23409852385520935}
+{"time": "06:54:13", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5731131434440613, "noul_T": 0.8532130995982341, "noul_auc": 0.7512615820830029, "noul_acc": 0.704, "noul_f1": 0.765079365079365, "noul_nll": 0.5736514215767383, "noul_ece": 0.052701245307922376, "noul_brier": 0.1954938496230899, "noul_sharp": 0.17823639512062073, "entropy": 0.5969312787055969}
+{"time": "06:54:56", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9113372, "tok_per_s": 3214.371023602123, "rows_per_s": 5.940329960631729, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5595800280570984, "noul_T": 0.8495371650607102, "choice_ce": 0.6842275261878967, "choice_T": 1.0560892190772713, "score_ce": 1.068756341934204, "score_T": 0.960112119036335, "noul_auc": 0.7447338666666667, "noul_acc": 0.695, "noul_f1": 0.7677075399847677, "noul_nll": 0.5717920997068286, "noul_ece": 0.04427964748442174, "noul_brier": 0.19495265463736913, "noul_sharp": 0.1802748143672943, "entropy": 0.5956453680992126}, "dropped_long": 0, "peak_gib": 18.284047603607178}

e28/S1-E28-E18A100-rt-syn/DONE

@@ -0,0 +1 @@
+2026-10-21 05:41:56

e28/S1-E28-E18A100-rt-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

e28/S1-E28-E18A100-rt-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",
+    "q_proj",
+    "k_proj",
+    "in_proj_qkv",
+    "in_proj_z",
+    "gate_proj",
+    "o_proj",
+    "up_proj",
+    "down_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
+}

e28/S1-E28-E18A100-rt-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfe4f81b3bbc62200aedfb4e41c5a27c4008f64325bd8f3696c58e61dea1d9a4
+size 160486456

e28/S1-E28-E18A100-rt-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
+}

e28/S1-E28-E18A100-rt-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:674988e4c1c9de8f20725c8477e8412211e6b386d03f787a68b365e10e0ecccb
+size 176641832

e28/S1-E28-E18A100-rt-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.0003491307725198567,
+  -0.0003493900876492262,
+  -0.0028010269161313772,
+  -0.001637146924622357,
+  -0.0005709006800316274,
+  0.001909777638502419,
+  0.0012821187265217304,
+  -0.002416141564026475,
+  -0.00029356408049352467,
+  -0.0006167300161905587,
+  -0.00018578222079668194,
+  0.0005398895591497421,
+  0.00042200679308734834,
+  -0.0036859940737485886,
+  -0.004552128259092569,
+  -0.00518615311011672,
+  -0.005845425650477409,
+  -0.004403236787766218,
+  -0.00495922239497304,
+  -0.005086651537567377,
+  -0.003894349792972207,
+  -0.0034869785886257887,
+  -0.005726017523556948,
+  -0.0035390020348131657
+ ],
+ "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"
+}

e28/S1-E28-E18A100-rt-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.8950098446145761,
+  "choice": 0.979539418518294,
+  "score": 0.9799331682889751
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.8950098446145761,
+   "ce_before": 0.5527104139328003,
+   "ce_after": 0.5517128109931946
+  },
+  "choice": {
+   "n": 60,
+   "T": 0.979539418518294,
+   "ce_before": 0.6627646684646606,
+   "ce_after": 0.6626709699630737
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9799331682889751,
+   "ce_before": 1.0634764432907104,
+   "ce_after": 1.0633518695831299
+  }
+ },
+ "source": "models/e28/S1-E28-E18A100-rt-syn/ckpt-16000"
+}

e28/S1-E28-E18A100-rt-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d934b88b444bb3cdf13f80e6ae05713ce4742c3c68bf0753ad84e6bbef715ffa
+size 393464

e28/S1-E28-E18A100-rt-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/e28/S1-E28-E18A100-rt-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e28/E28-E18A100-rt",
+  "data_manifest": {
+   "arm": "G-d512-n16k-a100",
+   "name": null,
+   "train_sha256": "e68e935b88943ba54ed4296ecedd5f5b36a7b9f114cd4d203c4f67b6ff70b912",
+   "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": 9113372,
+  "tok_per_s": 3211.5324866181877,
+  "rows_per_s": 5.935084194122323,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5527104139328003,
+   "noul_T": 0.8950098446145761,
+   "choice_ce": 0.6627646684646606,
+   "choice_T": 0.979539418518294,
+   "score_ce": 1.0634764432907104,
+   "score_T": 0.9799331682889751,
+   "noul_auc": 0.7475370666666667,
+   "noul_acc": 0.709,
+   "noul_f1": 0.7826736370425691,
+   "noul_nll": 0.5664274678044021,
+   "noul_ece": 0.02627825298905373,
+   "noul_brier": 0.19344517944405668,
+   "noul_sharp": 0.18980935215950012,
+   "entropy": 0.5791373252868652
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.284047603607178
+ }
+}

e28/S1-E28-E18A100-rt-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "04:54:38", "event": "start", "data": "data/e28/E28-E18A100-rt", "data_manifest": {"arm": "G-d512-n16k-a100", "name": null, "train_sha256": "e68e935b88943ba54ed4296ecedd5f5b36a7b9f114cd4d203c4f67b6ff70b912", "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": 4.754826784133911}
+{"time": "04:55:32", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 3324.3171051331515, "loss": 0.5821816921234131, "entropy": 0.5290134195238352, "sharp": 0.2557740639895201}
+{"time": "04:56:26", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 3342.619469628141, "loss": 0.6144347507506609, "entropy": 0.5189478620886803, "sharp": 0.2612189743667841}
+{"time": "04:57:21", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 3349.542457682998, "loss": 0.6007676869630814, "entropy": 0.6021645925939083, "sharp": 0.18828725535422564}
+{"time": "04:58:14", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 3355.5349822500643, "loss": 0.5568909235298634, "entropy": 0.6301036849617958, "sharp": 0.15472890809178352}
+{"time": "04:59:08", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 3366.9007224387783, "loss": 0.5048067662864923, "entropy": 0.5275742635130882, "sharp": 0.24421783722937107}
+{"time": "05:00:01", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 3373.9622074492613, "loss": 0.6004810445010662, "entropy": 0.6613579243421555, "sharp": 0.10816137958317995}
+{"time": "05:00:56", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 3370.7861903597727, "loss": 0.5786934606730938, "entropy": 0.5722468830645084, "sharp": 0.2008360568434}
+{"time": "05:01:50", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 3371.951818793317, "loss": 0.6301903324201703, "entropy": 0.5992131233215332, "sharp": 0.1787990666925907}
+{"time": "05:02:45", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 3376.198985771632, "loss": 0.6347234332934022, "entropy": 0.5524705238640308, "sharp": 0.22991735022515059}
+{"time": "05:03:39", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 3375.2733190336553, "loss": 0.5359861133620143, "entropy": 0.6075579188764095, "sharp": 0.17376595921814442}
+{"time": "05:03:59", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.5762508511543274, "noul_T": 0.8315131925978888, "noul_auc": 0.7390255016792964, "noul_acc": 0.698, "noul_f1": 0.7644305772230889, "noul_nll": 0.5822278542816639, "noul_ece": 0.04042269921302796, "noul_brier": 0.19934762721023885, "noul_sharp": 0.17159971594810486, "entropy": 0.6025501489639282}
+{"time": "05:04:52", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 3270.170828666978, "loss": 0.6028715297579765, "entropy": 0.5942024327814579, "sharp": 0.17740307934582233}
+{"time": "05:05:49", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 3282.816554544617, "loss": 0.5894755460321903, "entropy": 0.5931592807173729, "sharp": 0.17538204602897167}
+{"time": "05:06:45", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 3293.4199258042568, "loss": 0.571539293974638, "entropy": 0.5247620716691017, "sharp": 0.2440053764730692}
+{"time": "05:07:39", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 3299.1396210336193, "loss": 0.5822777207940817, "entropy": 0.5939646065235138, "sharp": 0.190550297498703}
+{"time": "05:08:34", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 3304.9054761201787, "loss": 0.6522177197039127, "entropy": 0.6408785283565521, "sharp": 0.1347259385511279}
+{"time": "05:09:27", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 3309.6455791802023, "loss": 0.6231949385255575, "entropy": 0.5634317472577095, "sharp": 0.1964956596493721}
+{"time": "05:10:21", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 3314.5241506812804, "loss": 0.5992662776261568, "entropy": 0.5705082695931196, "sharp": 0.20299063995480537}
+{"time": "05:11:15", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 3318.6283858023417, "loss": 0.5902198925614357, "entropy": 0.6148689202964306, "sharp": 0.15836940333247185}
+{"time": "05:12:10", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 3320.459810493686, "loss": 0.6276428215205669, "entropy": 0.552248153835535, "sharp": 0.21375497709959745}
+{"time": "05:13:04", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 3324.3459955267745, "loss": 0.5765422154217958, "entropy": 0.5640443600714207, "sharp": 0.21447045169770718}
+{"time": "05:13:24", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.5789938569068909, "noul_T": 1.0509989837525628, "noul_auc": 0.7417258780442523, "noul_acc": 0.68, "noul_f1": 0.7777777777777778, "noul_nll": 0.5867046617344022, "noul_ece": 0.0768196873664856, "noul_brier": 0.20149740134890642, "noul_sharp": 0.2089177817106247, "entropy": 0.5699007511138916}
+{"time": "05:14:19", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 3270.421274806653, "loss": 0.5589456427842379, "entropy": 0.580105422064662, "sharp": 0.2016209289431572}
+{"time": "05:15:13", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 3271.474330897922, "loss": 0.6095458008348942, "entropy": 0.5692577138543129, "sharp": 0.20771372877061367}
+{"time": "05:16:06", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 3275.132532625415, "loss": 0.5523778740316629, "entropy": 0.5647548958659172, "sharp": 0.22760403528809547}
+{"time": "05:16:59", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 3278.9549972163577, "loss": 0.5607727654278278, "entropy": 0.5328236278146505, "sharp": 0.23594816960394382}
+{"time": "05:17:57", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 3282.0149123861243, "loss": 0.6025251597166061, "entropy": 0.5898406133055687, "sharp": 0.18346295692026615}
+{"time": "05:18:48", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 3284.25290457319, "loss": 0.5806223824620247, "entropy": 0.5832313634455204, "sharp": 0.17723227478563786}
+{"time": "05:19:42", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 3288.008156761604, "loss": 0.6405539694242179, "entropy": 0.5990578662604094, "sharp": 0.17211627960205078}
+{"time": "05:20:35", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 3291.6077544568147, "loss": 0.6058562472462654, "entropy": 0.539489358663559, "sharp": 0.2305415105074644}
+{"time": "05:21:30", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 3294.54182375482, "loss": 0.599469605833292, "entropy": 0.5510324835777283, "sharp": 0.21221148874610662}
+{"time": "05:22:25", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 3298.2929114311473, "loss": 0.5266063790768385, "entropy": 0.576934652402997, "sharp": 0.17126601748168468}
+{"time": "05:22:44", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.57090163230896, "noul_T": 0.7756709156364596, "noul_auc": 0.7441224620681508, "noul_acc": 0.7, "noul_f1": 0.7692307692307693, "noul_nll": 0.5765361273288727, "noul_ece": 0.06336354553699494, "noul_brier": 0.19727589353971983, "noul_sharp": 0.1659199595451355, "entropy": 0.6051366925239563}
+{"time": "05:23:37", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 3262.4900043871453, "loss": 0.5625473354011774, "entropy": 0.5803492031991482, "sharp": 0.18244618922472}
+{"time": "05:24:28", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 3265.229639956781, "loss": 0.6385385170578957, "entropy": 0.6035343185067177, "sharp": 0.16779274493455887}
+{"time": "05:25:22", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 3269.351559759809, "loss": 0.5611035637557507, "entropy": 0.5464083477854729, "sharp": 0.22834619507193565}
+{"time": "05:26:15", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 3272.9586242924706, "loss": 0.6386581789702177, "entropy": 0.6003948114812374, "sharp": 0.167050426825881}
+{"time": "05:27:07", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 3275.8133966874448, "loss": 0.6118316054344177, "entropy": 0.5887851566076279, "sharp": 0.17444227263331413}
+{"time": "05:28:00", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 3279.414812985462, "loss": 0.6040074042975903, "entropy": 0.6173115968704224, "sharp": 0.16369462944567204}
+{"time": "05:28:52", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 3283.0091197545535, "loss": 0.6168727185577154, "entropy": 0.5541434902697802, "sharp": 0.21848703734576702}
+{"time": "05:29:48", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 3285.94642100366, "loss": 0.5934649743139744, "entropy": 0.5759682320058346, "sharp": 0.19066925160586834}
+{"time": "05:30:43", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 3289.937870578109, "loss": 0.626618885435164, "entropy": 0.6006878651678562, "sharp": 0.1617059651762247}
+{"time": "05:31:35", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 3291.229789227373, "loss": 0.5811722055077553, "entropy": 0.5933747291564941, "sharp": 0.1702809166163206}
+{"time": "05:31:55", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.5681905746459961, "noul_T": 0.9391060152874654, "noul_auc": 0.7450338390913234, "noul_acc": 0.698, "noul_f1": 0.7722473604826546, "noul_nll": 0.5738915740866214, "noul_ece": 0.035751788198947906, "noul_brier": 0.19638482748407443, "noul_sharp": 0.1894688606262207, "entropy": 0.5781760811805725}
+{"time": "05:32:48", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 3264.4773228562294, "loss": 0.5989872589707375, "entropy": 0.5823234394192696, "sharp": 0.19112121406942606}
+{"time": "05:33:39", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 3266.3483477236264, "loss": 0.6516177747398615, "entropy": 0.6142925452440977, "sharp": 0.1551995575428009}
+{"time": "05:34:34", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 3269.2346552535387, "loss": 0.5770297460258007, "entropy": 0.5725030777975917, "sharp": 0.18772407714277506}
+{"time": "05:35:27", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 3271.4504029767604, "loss": 0.5983109809458256, "entropy": 0.5960347801446915, "sharp": 0.1682441271841526}
+{"time": "05:36:22", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 3273.740264407381, "loss": 0.5502499658614397, "entropy": 0.5791618190705776, "sharp": 0.19975292682647705}
+{"time": "05:37:14", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 3275.4269044173816, "loss": 0.5780162587761879, "entropy": 0.5901758074760437, "sharp": 0.19233804941177368}
+{"time": "05:38:06", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 3277.5185757402824, "loss": 0.5851897373795509, "entropy": 0.5742927268147469, "sharp": 0.20923052355647087}
+{"time": "05:39:02", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 3279.550906167545, "loss": 0.5837513078004122, "entropy": 0.5576575566083193, "sharp": 0.22102190181612968}
+{"time": "05:39:57", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 3281.5783665266927, "loss": 0.6089657805860043, "entropy": 0.5673855915665627, "sharp": 0.20276403427124023}
+{"time": "05:40:53", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 3283.682019469101, "loss": 0.5507879704236984, "entropy": 0.5276962127536535, "sharp": 0.24716411530971527}
+{"time": "05:41:13", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.5671284198760986, "noul_T": 0.915414303765002, "noul_auc": 0.7490337715819142, "noul_acc": 0.71, "noul_f1": 0.7744945567651633, "noul_nll": 0.5708517383821309, "noul_ece": 0.037796863555908215, "noul_brier": 0.19516738459980418, "noul_sharp": 0.1869892030954361, "entropy": 0.5807954668998718}
+{"time": "05:41:56", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 9113372, "tok_per_s": 3211.5324866181877, "rows_per_s": 5.935084194122323, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5527104139328003, "noul_T": 0.8950098446145761, "choice_ce": 0.6627646684646606, "choice_T": 0.979539418518294, "score_ce": 1.0634764432907104, "score_T": 0.9799331682889751, "noul_auc": 0.7475370666666667, "noul_acc": 0.709, "noul_f1": 0.7826736370425691, "noul_nll": 0.5664274678044021, "noul_ece": 0.02627825298905373, "noul_brier": 0.19344517944405668, "noul_sharp": 0.18980935215950012, "entropy": 0.5791373252868652}, "dropped_long": 0, "peak_gib": 18.284047603607178}

e28/S1-E28-E30-a1-rw-syn/DONE

@@ -0,0 +1 @@
+2026-10-21 18:40:28

e28/S1-E28-E30-a1-rw-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

e28/S1-E28-E30-a1-rw-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",
+    "gate_proj",
+    "in_proj_z",
+    "up_proj",
+    "o_proj",
+    "q_proj",
+    "k_proj",
+    "down_proj",
+    "in_proj_qkv"
+  ],
+  "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
+}

e28/S1-E28-E30-a1-rw-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d78eef6a2ecf5d7a7fa2e3e767d90b6a47985a8b9f58e1394629c81783d1bcc5
+size 160486456

e28/S1-E28-E30-a1-rw-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
+}

e28/S1-E28-E30-a1-rw-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f90749a62d8e87daafce65f9bd5e2326e6c48bd746b4755d9e51db4bb4547bc
+size 176641832

e28/S1-E28-E30-a1-rw-syn/ckpt-16000/adapter_vllm/decision_head.json

@@ -0,0 +1,99 @@
+{
+ "bias": [
+  0.0005390653386712074,
+  -0.0005393188330344856,
+  -0.002505986951291561,
+  -0.0016197734512388706,
+  -0.0006678372155874968,
+  0.0021317743230611086,
+  0.001403430593200028,
+  -0.002808196237310767,
+  -0.0001560490927658975,
+  -0.0002977506082970649,
+  0.0003498705627862364,
+  0.0007481754291802645,
+  -0.0015342392725870013,
+  -0.0035954874474555254,
+  -0.004323368426412344,
+  -0.0050967843271791935,
+  -0.005942096468061209,
+  -0.004360170569270849,
+  -0.0049414122477173805,
+  -0.00499636260792613,
+  -0.003998260945081711,
+  -0.00342180160805583,
+  -0.005643993150442839,
+  -0.0035076222848147154
+ ],
+ "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"
+}

e28/S1-E28-E30-a1-rw-syn/ckpt-16000/calibration.json

@@ -0,0 +1,29 @@
+{
+ "version": 1,
+ "per_kind": {
+  "noul": 0.7893042861110305,
+  "choice": 1.1089171735549759,
+  "score": 0.9725362941237807
+ },
+ "fit": {
+  "noul": {
+   "n": 1087,
+   "T": 0.7893042861110305,
+   "ce_before": 0.5852800011634827,
+   "ce_after": 0.5815325975418091
+  },
+  "choice": {
+   "n": 60,
+   "T": 1.1089171735549759,
+   "ce_before": 0.6824166774749756,
+   "ce_after": 0.6801118850708008
+  },
+  "score": {
+   "n": 53,
+   "T": 0.9725362941237807,
+   "ce_before": 1.0658724308013916,
+   "ce_after": 1.065628170967102
+  }
+ },
+ "source": "models/e28/S1-E28-E30-a1-rw-syn/ckpt-16000"
+}

e28/S1-E28-E30-a1-rw-syn/ckpt-16000/head.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e188d1d84a9fa6d952cdd6536db406217276ece45e56249448536c47c9d523fd
+size 393464

e28/S1-E28-E30-a1-rw-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/e28/S1-E28-E30-a1-rw-syn/ckpt-16000",
+ "softcap": null,
+ "e16": {
+  "data": "data/e28/E28-E30-a1-rw",
+  "data_manifest": {
+   "arm": "E28-E30-a1-rw",
+   "name": null,
+   "train_sha256": "d26b35d46e3927d8807524aa9b0efcfcb8505a3ac3b2c30a9a2933198ec3e62d",
+   "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": 8988859,
+  "tok_per_s": 3275.2473404806465,
+  "rows_per_s": 6.13667604304651,
+  "val": {
+   "n": 1200,
+   "dropped_long": 0,
+   "noul_ce": 0.5852800011634827,
+   "noul_T": 0.7893042861110305,
+   "choice_ce": 0.6824166774749756,
+   "choice_T": 1.1089171735549759,
+   "score_ce": 1.0658724308013916,
+   "score_T": 0.9725362941237807,
+   "noul_auc": 0.7184704,
+   "noul_acc": 0.696,
+   "noul_f1": 0.76793893129771,
+   "noul_nll": 0.5946063060313463,
+   "noul_ece": 0.059608851045370115,
+   "noul_brier": 0.20349168514643015,
+   "noul_sharp": 0.15316306054592133,
+   "entropy": 0.6263568997383118
+  },
+  "dropped_long": 0,
+  "peak_gib": 18.282336711883545
+ }
+}

e28/S1-E28-E30-a1-rw-syn/train_log.jsonl

@@ -0,0 +1,57 @@
+{"time": "17:54:43", "event": "start", "data": "data/e28/E28-E30-a1-rw", "data_manifest": {"arm": "E28-E30-a1-rw", "name": null, "train_sha256": "d26b35d46e3927d8807524aa9b0efcfcb8505a3ac3b2c30a9a2933198ec3e62d", "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": 4.869741678237915}
+{"time": "17:55:35", "event": "step", "step": 10, "rows_seen": 320, "lr": 7.333333333333333e-05, "tok_per_s": 3377.2851167766844, "loss": 0.6338330507278442, "entropy": 0.6497155390679836, "sharp": 0.12389896251261234}
+{"time": "17:56:26", "event": "step", "step": 20, "rows_seen": 640, "lr": 9.997508952966197e-05, "tok_per_s": 3412.0501605520712, "loss": 0.6315236687660217, "entropy": 0.645004142075777, "sharp": 0.1294181663542986}
+{"time": "17:57:18", "event": "step", "step": 30, "rows_seen": 960, "lr": 9.977596250541417e-05, "tok_per_s": 3426.6409228336306, "loss": 0.7279861196875572, "entropy": 0.6225987821817398, "sharp": 0.14857110660523176}
+{"time": "17:58:11", "event": "step", "step": 40, "rows_seen": 1280, "lr": 9.937854366485982e-05, "tok_per_s": 3432.445975142808, "loss": 0.6418519904837012, "entropy": 0.6383919306099415, "sharp": 0.15334222465753555}
+{"time": "17:59:01", "event": "step", "step": 50, "rows_seen": 1600, "lr": 9.8784499920729e-05, "tok_per_s": 3446.264373239778, "loss": 0.6650119200348854, "entropy": 0.669834554195404, "sharp": 0.08912533335387707}
+{"time": "17:59:53", "event": "step", "step": 60, "rows_seen": 1920, "lr": 9.79963228989294e-05, "tok_per_s": 3449.69168789776, "loss": 0.6345913931727409, "entropy": 0.661293052136898, "sharp": 0.10877801477909088}
+{"time": "18:00:46", "event": "step", "step": 70, "rows_seen": 2240, "lr": 9.701731848780134e-05, "tok_per_s": 3447.4641306971735, "loss": 0.6558307074010372, "entropy": 0.6163828521966934, "sharp": 0.17564489413052797}
+{"time": "18:01:38", "event": "step", "step": 80, "rows_seen": 2560, "lr": 9.585159297207315e-05, "tok_per_s": 3446.495126628318, "loss": 0.6458177231252193, "entropy": 0.6574584618210793, "sharp": 0.09814455173909664}
+{"time": "18:02:30", "event": "step", "step": 90, "rows_seen": 2880, "lr": 9.450403580967612e-05, "tok_per_s": 3447.1536995850843, "loss": 0.7036329247057438, "entropy": 0.6312954239547253, "sharp": 0.13076926954090595}
+{"time": "18:03:21", "event": "step", "step": 100, "rows_seen": 3200, "lr": 9.298029912365855e-05, "tok_per_s": 3443.790074357248, "loss": 0.6426000259816647, "entropy": 0.6454283036291599, "sharp": 0.12996309995651245}
+{"time": "18:03:41", "event": "val", "step": 100, "rows_seen": 3200, "n": 500, "dropped_long": 0, "noul_ce": 0.6357130408287048, "noul_T": 0.6211001554924073, "noul_auc": 0.7169161701912204, "noul_acc": 0.67, "noul_f1": 0.7169811320754716, "noul_nll": 0.6331049682497978, "noul_ece": 0.12251776278018953, "noul_brier": 0.22072263777458637, "noul_sharp": 0.0979577973484993, "entropy": 0.66475510597229}
+{"time": "18:04:33", "event": "step", "step": 110, "rows_seen": 3520, "lr": 9.128677399521752e-05, "tok_per_s": 3329.341525297175, "loss": 0.5837785638868809, "entropy": 0.594234935939312, "sharp": 0.17981287464499474}
+{"time": "18:05:24", "event": "step", "step": 120, "rows_seen": 3840, "lr": 8.943056365728335e-05, "tok_per_s": 3340.8014945412456, "loss": 0.6435245983302593, "entropy": 0.6653639450669289, "sharp": 0.08566387463361025}
+{"time": "18:06:16", "event": "step", "step": 130, "rows_seen": 4160, "lr": 8.741945370109202e-05, "tok_per_s": 3349.8566168747334, "loss": 0.5971063524484634, "entropy": 0.6157004665583372, "sharp": 0.1654625367373228}
+{"time": "18:07:07", "event": "step", "step": 140, "rows_seen": 4480, "lr": 8.526187942070997e-05, "tok_per_s": 3355.556275132598, "loss": 0.6818988248705864, "entropy": 0.630051501095295, "sharp": 0.1473753023892641}
+{"time": "18:07:59", "event": "step", "step": 150, "rows_seen": 4800, "lr": 8.296689043247915e-05, "tok_per_s": 3360.7227262118395, "loss": 0.5924976449459791, "entropy": 0.6424582377076149, "sharp": 0.13546935934573412}
+{"time": "18:08:49", "event": "step", "step": 160, "rows_seen": 5120, "lr": 8.054411271778132e-05, "tok_per_s": 3365.882176538322, "loss": 0.7351751737296581, "entropy": 0.6749421320855618, "sharp": 0.08052086178213358}
+{"time": "18:09:42", "event": "step", "step": 170, "rows_seen": 5440, "lr": 7.800370824832674e-05, "tok_per_s": 3372.250622334071, "loss": 0.627573162317276, "entropy": 0.6558959558606148, "sharp": 0.12126355338841677}
+{"time": "18:10:34", "event": "step", "step": 180, "rows_seen": 5760, "lr": 7.535633236331324e-05, "tok_per_s": 3374.350063685714, "loss": 0.5482108443975449, "entropy": 0.6002820134162903, "sharp": 0.15795409120619297}
+{"time": "18:11:26", "event": "step", "step": 190, "rows_seen": 6080, "lr": 7.26130890772308e-05, "tok_per_s": 3378.7029186270856, "loss": 0.6277678199112415, "entropy": 0.6315691769123077, "sharp": 0.1291743665933609}
+{"time": "18:12:20", "event": "step", "step": 200, "rows_seen": 6400, "lr": 6.978548450576568e-05, "tok_per_s": 3386.2736576786465, "loss": 0.5901587307453156, "entropy": 0.640953928232193, "sharp": 0.12728234101086855}
+{"time": "18:12:40", "event": "val", "step": 200, "rows_seen": 6400, "n": 500, "dropped_long": 0, "noul_ce": 0.6395695209503174, "noul_T": 0.5691573306075144, "noul_auc": 0.7214393006025215, "noul_acc": 0.658, "noul_f1": 0.7005253940455342, "noul_nll": 0.6360012717843055, "noul_ece": 0.14528395158052446, "noul_brier": 0.2222243668546396, "noul_sharp": 0.08701810240745544, "entropy": 0.6687576770782471}
+{"time": "18:13:32", "event": "step", "step": 210, "rows_seen": 6720, "lr": 6.688537860515266e-05, "tok_per_s": 3328.7936073683777, "loss": 0.6967102643102407, "entropy": 0.6456865444779396, "sharp": 0.11548532731831074}
+{"time": "18:14:23", "event": "step", "step": 220, "rows_seen": 7040, "lr": 6.392493542739706e-05, "tok_per_s": 3333.543467474346, "loss": 0.6308403629809618, "entropy": 0.6531947795301676, "sharp": 0.10847062431275845}
+{"time": "18:15:16", "event": "step", "step": 230, "rows_seen": 7360, "lr": 6.091657210001309e-05, "tok_per_s": 3340.0488545184285, "loss": 0.6255427449941635, "entropy": 0.6058772392570972, "sharp": 0.16990910656750202}
+{"time": "18:16:10", "event": "step", "step": 240, "rows_seen": 7680, "lr": 5.7872906744276015e-05, "tok_per_s": 3344.300160322416, "loss": 0.71213773265481, "entropy": 0.6475676037371159, "sharp": 0.12428359594196081}
+{"time": "18:17:00", "event": "step", "step": 250, "rows_seen": 8000, "lr": 5.48067055504363e-05, "tok_per_s": 3348.6065993451734, "loss": 0.6979744881391525, "entropy": 0.6062525287270546, "sharp": 0.18239591270685196}
+{"time": "18:17:53", "event": "step", "step": 260, "rows_seen": 8320, "lr": 5.173082923188119e-05, "tok_per_s": 3352.879972166113, "loss": 0.5990705713629723, "entropy": 0.6486502215266228, "sharp": 0.1148771271109581}
+{"time": "18:18:44", "event": "step", "step": 270, "rows_seen": 8640, "lr": 4.865817908283331e-05, "tok_per_s": 3358.1137566191123, "loss": 0.6887011472135782, "entropy": 0.6524376422166824, "sharp": 0.1245900746434927}
+{"time": "18:19:38", "event": "step", "step": 280, "rows_seen": 8960, "lr": 4.560164286583922e-05, "tok_per_s": 3363.13589939421, "loss": 0.6507637053728104, "entropy": 0.6164841204881668, "sharp": 0.16536646895110607}
+{"time": "18:20:30", "event": "step", "step": 290, "rows_seen": 9280, "lr": 4.257404075601443e-05, "tok_per_s": 3366.5677356122173, "loss": 0.6710220128297806, "entropy": 0.6520177498459816, "sharp": 0.11389991734176874}
+{"time": "18:21:22", "event": "step", "step": 300, "rows_seen": 9600, "lr": 3.958807156877371e-05, "tok_per_s": 3368.685068466974, "loss": 0.6416098596528172, "entropy": 0.6477503851056099, "sharp": 0.12916356604546309}
+{"time": "18:21:42", "event": "val", "step": 300, "rows_seen": 9600, "n": 500, "dropped_long": 0, "noul_ce": 0.6175177693367004, "noul_T": 0.7215910214407419, "noul_auc": 0.7286628073787784, "noul_acc": 0.674, "noul_f1": 0.7165217391304348, "noul_nll": 0.6141112993955612, "noul_ece": 0.11070850026607516, "noul_brier": 0.21205119923383237, "noul_sharp": 0.12754103541374207, "entropy": 0.6471936702728271}
+{"time": "18:22:35", "event": "step", "step": 310, "rows_seen": 9920, "lr": 3.665625949658603e-05, "tok_per_s": 3331.3107669826086, "loss": 0.6511296331882477, "entropy": 0.6407931558787823, "sharp": 0.13699984084814787}
+{"time": "18:23:27", "event": "step", "step": 320, "rows_seen": 10240, "lr": 3.379090157815887e-05, "tok_per_s": 3334.268322375747, "loss": 0.6851719468832016, "entropy": 0.6582037210464478, "sharp": 0.11040528025478125}
+{"time": "18:24:20", "event": "step", "step": 330, "rows_seen": 10560, "lr": 3.10040161203844e-05, "tok_per_s": 3338.41235718141, "loss": 0.6316585764288902, "entropy": 0.6686232909560204, "sharp": 0.09014521725475788}
+{"time": "18:25:12", "event": "step", "step": 340, "rows_seen": 10880, "lr": 2.8307292289383776e-05, "tok_per_s": 3341.2341405458083, "loss": 0.6132487095892429, "entropy": 0.5950825084000826, "sharp": 0.18371212854981422}
+{"time": "18:26:04", "event": "step", "step": 350, "rows_seen": 11200, "lr": 2.571204108208194e-05, "tok_per_s": 3344.3896086410027, "loss": 0.669573157094419, "entropy": 0.6661554910242558, "sharp": 0.09097274485975504}
+{"time": "18:26:56", "event": "step", "step": 360, "rows_seen": 11520, "lr": 2.322914788395595e-05, "tok_per_s": 3348.0805313321516, "loss": 0.6744805574417114, "entropy": 0.6449346467852592, "sharp": 0.12108333129435778}
+{"time": "18:27:49", "event": "step", "step": 370, "rows_seen": 11840, "lr": 2.0869026811945157e-05, "tok_per_s": 3350.44983143245, "loss": 0.6190249249339104, "entropy": 0.5869340002536774, "sharp": 0.19311721064150333}
+{"time": "18:28:42", "event": "step", "step": 380, "rows_seen": 12160, "lr": 1.8641577034025627e-05, "tok_per_s": 3352.2686568872564, "loss": 0.6518426295369864, "entropy": 0.6288196444511414, "sharp": 0.13674758467823267}
+{"time": "18:29:33", "event": "step", "step": 390, "rows_seen": 12480, "lr": 1.6556141248659472e-05, "tok_per_s": 3355.277724566154, "loss": 0.697613812983036, "entropy": 0.637400783598423, "sharp": 0.12889212369918823}
+{"time": "18:30:25", "event": "step", "step": 400, "rows_seen": 12800, "lr": 1.4621466498270773e-05, "tok_per_s": 3356.9764095931605, "loss": 0.61271165125072, "entropy": 0.6158593855798244, "sharp": 0.1538755875080824}
+{"time": "18:30:45", "event": "val", "step": 400, "rows_seen": 12800, "n": 500, "dropped_long": 0, "noul_ce": 0.6081778407096863, "noul_T": 0.7508601228503166, "noul_auc": 0.7353631162343252, "noul_acc": 0.706, "noul_f1": 0.7578253706754531, "noul_nll": 0.6025428766608238, "noul_ece": 0.10199650961160661, "noul_brier": 0.20667334150308625, "noul_sharp": 0.13835392892360687, "entropy": 0.63767409324646}
+{"time": "18:31:38", "event": "step", "step": 410, "rows_seen": 13120, "lr": 1.284566748110963e-05, "tok_per_s": 3329.4294478765714, "loss": 0.6279964409768581, "entropy": 0.6595849394798279, "sharp": 0.0933420555666089}
+{"time": "18:32:30", "event": "step", "step": 420, "rows_seen": 13440, "lr": 1.1236192515387821e-05, "tok_per_s": 3332.267656872563, "loss": 0.6110328286886215, "entropy": 0.6364967674016953, "sharp": 0.13975976780056953}
+{"time": "18:33:21", "event": "step", "step": 430, "rows_seen": 13760, "lr": 9.799792298443728e-06, "tok_per_s": 3334.403992239488, "loss": 0.6370786568149924, "entropy": 0.6099487831816077, "sharp": 0.15920150559395552}
+{"time": "18:34:15", "event": "step", "step": 440, "rows_seen": 14080, "lr": 8.542491591971554e-06, "tok_per_s": 3338.3739954868947, "loss": 0.6492991335690022, "entropy": 0.6316069588065147, "sharp": 0.13702389132231474}
+{"time": "18:35:07", "event": "step", "step": 450, "rows_seen": 14400, "lr": 7.469563952077155e-06, "tok_per_s": 3340.5295980148544, "loss": 0.5848479829728603, "entropy": 0.6442913822829723, "sharp": 0.11933630052953959}
+{"time": "18:35:58", "event": "step", "step": 460, "rows_seen": 14720, "lr": 6.585509610150972e-06, "tok_per_s": 3342.722650089894, "loss": 0.6680366033688188, "entropy": 0.6582284085452557, "sharp": 0.10449005477130413}
+{"time": "18:36:50", "event": "step", "step": 470, "rows_seen": 15040, "lr": 5.894036597333646e-06, "tok_per_s": 3344.5248878335597, "loss": 0.6666250005364418, "entropy": 0.6220298185944557, "sharp": 0.15243109688162804}
+{"time": "18:37:42", "event": "step", "step": 480, "rows_seen": 15360, "lr": 5.398045191744809e-06, "tok_per_s": 3346.176169703262, "loss": 0.6240421757102013, "entropy": 0.6295576170086861, "sharp": 0.13606482464820147}
+{"time": "18:38:33", "event": "step", "step": 490, "rows_seen": 15680, "lr": 5.099615753708697e-06, "tok_per_s": 3349.000266115592, "loss": 0.6278671249747276, "entropy": 0.6624680310487747, "sharp": 0.09166452661156654}
+{"time": "18:39:25", "event": "step", "step": 500, "rows_seen": 16000, "lr": 5e-06, "tok_per_s": 3351.4645199172123, "loss": 0.7027888223528862, "entropy": 0.6221847385168076, "sharp": 0.15409976709634066}
+{"time": "18:39:45", "event": "val", "step": 500, "rows_seen": 16000, "n": 500, "dropped_long": 0, "noul_ce": 0.6007530093193054, "noul_T": 0.7747629570244622, "noul_auc": 0.7332196924946415, "noul_acc": 0.708, "noul_f1": 0.7660256410256411, "noul_nll": 0.5951983838230371, "noul_ece": 0.08512620198726653, "noul_brier": 0.20347204683407666, "noul_sharp": 0.14822502434253693, "entropy": 0.6300143599510193}
+{"time": "18:40:28", "event": "ckpt", "tag": "ckpt-16000", "rows_seen": 16000, "replay_seen": 842, "step": 500, "tokens": 8988859, "tok_per_s": 3275.2473404806465, "rows_per_s": 6.13667604304651, "val": {"n": 1200, "dropped_long": 0, "noul_ce": 0.5852800011634827, "noul_T": 0.7893042861110305, "choice_ce": 0.6824166774749756, "choice_T": 1.1089171735549759, "score_ce": 1.0658724308013916, "score_T": 0.9725362941237807, "noul_auc": 0.7184704, "noul_acc": 0.696, "noul_f1": 0.76793893129771, "noul_nll": 0.5946063060313463, "noul_ece": 0.059608851045370115, "noul_brier": 0.20349168514643015, "noul_sharp": 0.15316306054592133, "entropy": 0.6263568997383118}, "dropped_long": 0, "peak_gib": 18.282336711883545}

e28/S1-E28-E30-a1-uni-syn/DONE

@@ -0,0 +1 @@
+2026-10-21 17:27:13

e28/S1-E28-E30-a1-uni-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

e28/S1-E28-E30-a1-uni-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",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "gate_proj",
+    "in_proj_qkv",
+    "down_proj",
+    "q_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
+}

e28/S1-E28-E30-a1-uni-syn/ckpt-16000/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0bc376b8094b519732e128bc6c2e77088496bc2b7ef13135e140ee0912c34be
+size 160486456

e28/S1-E28-E30-a1-uni-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
+}

e28/S1-E28-E30-a1-uni-syn/ckpt-16000/adapter_vllm/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd954b72350ecb7f709fe16b769604f928bdb583b488ca6aa79d7f8adf78ca5b
+size 176641832