adgomant/adele-judge-qwen3-14B-cre
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: adgomant/adele-judge-qwen3-14B-cre
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-10-02 07:09:18 +08:00
commit cd82c37be3
25 changed files with 2809 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
.gitattributes vendored Normal file
View File

@@ -0,0 +1,37 @@
*.7z filter=lfs diff=lfs merge=lfs -text
*.arrow filter=lfs diff=lfs merge=lfs -text
*.bin filter=lfs diff=lfs merge=lfs -text
*.bz2 filter=lfs diff=lfs merge=lfs -text
*.ckpt filter=lfs diff=lfs merge=lfs -text
*.ftz filter=lfs diff=lfs merge=lfs -text
*.gz filter=lfs diff=lfs merge=lfs -text
*.h5 filter=lfs diff=lfs merge=lfs -text
*.joblib filter=lfs diff=lfs merge=lfs -text
*.lfs.* filter=lfs diff=lfs merge=lfs -text
*.mlmodel filter=lfs diff=lfs merge=lfs -text
*.model filter=lfs diff=lfs merge=lfs -text
*.msgpack filter=lfs diff=lfs merge=lfs -text
*.npy filter=lfs diff=lfs merge=lfs -text
*.npz filter=lfs diff=lfs merge=lfs -text
*.onnx filter=lfs diff=lfs merge=lfs -text
*.ot filter=lfs diff=lfs merge=lfs -text
*.parquet filter=lfs diff=lfs merge=lfs -text
*.pb filter=lfs diff=lfs merge=lfs -text
*.pickle filter=lfs diff=lfs merge=lfs -text
*.pkl filter=lfs diff=lfs merge=lfs -text
*.pt filter=lfs diff=lfs merge=lfs -text
*.pth filter=lfs diff=lfs merge=lfs -text
*.rar filter=lfs diff=lfs merge=lfs -text
*.safetensors filter=lfs diff=lfs merge=lfs -text
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*.tar.* filter=lfs diff=lfs merge=lfs -text
*.tar filter=lfs diff=lfs merge=lfs -text
*.tflite filter=lfs diff=lfs merge=lfs -text
*.tgz filter=lfs diff=lfs merge=lfs -text
*.wasm filter=lfs diff=lfs merge=lfs -text
*.xz filter=lfs diff=lfs merge=lfs -text
*.zip filter=lfs diff=lfs merge=lfs -text
*.zst filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text
adapter/tokenizer.json filter=lfs diff=lfs merge=lfs -text
tokenizer.json filter=lfs diff=lfs merge=lfs -text

151
README.md Normal file
View File

@@ -0,0 +1,151 @@
---
library_name: transformers
tags:
- text-generation
- peft
- adele
- judge
base_model: Qwen/Qwen3-14B
datasets:
- CFI-Kinds-of-Intelligence/ADeLe_battery_v1dot0
---
# ADeLe Distilled Judge
This repository contains an ADeLe-suite-specific distilled judge. It scores a model response against a question and reference answer with an ordinal score from 1 to 5, then derives binary correctness with the ADeLe threshold.
The repository root contains a merged Transformers model for standard loading. The original LoRA adapter is also included under `adapter/` for provenance and reuse.
## Intended Use
Use this model to score ADeLe-style examples where a question, reference answer, and model response are available. It is intended for out-of-model evaluation within the ADeLe benchmark suite, not as a general-purpose evaluator.
## Input Format
The recommended helper accepts:
- `question`
- `reference_answer` or `ground_truth`
- `model_response`
## Score Rubric
Allowed scores: 1, 2, 3, 4, 5
- 1: surely incorrect
- 2: likely incorrect
- 3: minimally correct or sufficient
- 4: likely correct
- 5: surely correct
Binary label: scores greater than or equal to 3 are `CORRECT`; lower scores are `INCORRECT`.
## Training And Validation Data
| Split | Examples | Models |
| --- | --- | --- |
| train | 239,420 | 16 |
| validation | 45,738 | 3 |
- `train` models: `DK-R1-Dist-Qwen-1.5B`, `DK-R1-Dist-Qwen-32B`, `DK-R1-Dist-Qwen-7B`, `gemini-2.5-flash`, `gemini-3.1-pro`, `gpt-35-turbo`, `gpt-5.2`, `gpt4o`, `llama3d1-405b`, `llama3d2-11b`, `llama3d2-1b`, `llama3d2-90b`, `llama4-17B-128E`, `o1-mini`, `o1_re=low`, `o3-mini`
- `validation` models: `DK-R1-Dist-Qwen-14B`, `gemini-3-flash`, `llama3d2-3b`
## Data Quality And Label Construction
Training labels are distilled from two proprietary judge scores used by the ADeLe evaluation pipeline to derive the official correctness signal. The configured source columns are `score_gpt4o` and `score_sonnet`.
- Ordinal target: `floor(mean(score_gpt4o, score_sonnet))`.
- Binary target: `CORRECT` when the ordinal target is >= `3`.
- Judge-agreement filter: keep examples with `abs(score_gpt4o - score_sonnet) <= 1`.
- Response-length filter: keep responses with at most `4096` base-tokenizer tokens before prompt formatting.
- Sequence-length filter: keep full chat-formatted examples within `max_seq_length=8192`.
## Validation Results
Source artifact: `validation_trainer_metrics.json`.
| Metric | Value |
| --- | --- |
| Epoch | 1.0000 |
| Binary accuracy | 0.9894 |
| Binary macro F1 | 0.9880 |
| Precision, CORRECT | 0.9932 |
| Recall, CORRECT | 0.9909 |
| Precision, INCORRECT | 0.9817 |
| Recall, INCORRECT | 0.9863 |
| False negative rate, CORRECT | 0.0091 |
| False positive rate, CORRECT | 0.0137 |
| Ordinal accuracy | 0.9639 |
| Ordinal macro F1 | 0.7351 |
| Mean confidence | 0.9604 |
## Recommended Inference
Do not use free-form generation as the primary prediction method. The recommended path scores the restricted continuations `"1"`, `"2"`, `"3"`, `"4"`, and `"5"`.
```python
from transformers import pipeline
judge = pipeline(
"adele-judge",
model="adgomant/adele-judge-qwen3-14-cre",
trust_remote_code=True,
device_map="auto",
)
result = judge(
{"question": "...", "reference_answer": "...", "model_response": "..."}
)
print(result)
results = judge([
{"question": "...", "reference_answer": "...", "model_response": "..."},
{"question": "...", "ground_truth": "...", "model_response": "..."},
], batch_size=8)
```
The result has this shape:
```python
{
"score": 4,
"label": "CORRECT",
"probs": {"1": 0.01, "2": 0.02, "3": 0.08, "4": 0.70, "5": 0.19},
"logprobs": {"1": -5.0, "2": -4.2, "3": -2.9, "4": -0.8, "5": -2.1},
"confidence": 0.70,
"margin": 1.3,
"entropy": 0.82,
}
```
## Standard Transformers Loading
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("adgomant/adele-judge-qwen3-14-cre", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("adgomant/adele-judge-qwen3-14-cre", trust_remote_code=True)
```
`generation_config.json` uses safe one-token defaults for debugging, but `generate()` is not the recommended scoring method.
## Metadata
Training, filtering, split, tokenization, and metric artifacts available at packaging time are stored in `adele_judge_metadata.json`.
The model is trained on distilled judge targets. These targets are useful for reproducing the ADeLe paper-style correctness signal at lower inference cost, but they should not be interpreted as independent human annotations.
## References
- ADeLe project page: [ADeLe v1.0](https://kinds-of-intelligence-cfi.github.io/ADELE/).
- ADeLe paper and official correctness definition: [General scales unlock AI evaluation with explanatory and predictive power](https://www.nature.com/articles/s41586-026-10303-2).
- Official ADeLe dataset: [CFI-Kinds-of-Intelligence/ADeLe_battery_v1dot0](https://huggingface.co/datasets/CFI-Kinds-of-Intelligence/ADeLe_battery_v1dot0).
- Official instance-level model-response data used for distillation: [https://github.com/Kinds-of-Intelligence-CFI/ADeLe-AIEvaluation/tree/main/ADeLe_battery_data/subject_specific_instance_level_data](https://github.com/Kinds-of-Intelligence-CFI/ADeLe-AIEvaluation/tree/main/ADeLe_battery_data/subject_specific_instance_level_data).
- Training and Hub packaging implementation: [https://github.com/adgomant/adele-judge](https://github.com/adgomant/adele-judge).
## Limitations
- ADeLe-specific judge; not a general-purpose evaluator.
- Distilled from proprietary judge labels and inherits their noise, calibration, and biases.
- Intended for scoring responses against a reference answer.
- It should not produce explanations; the expected output is a single score.
- Validation is out-of-model within the ADeLe suite, so transfer outside that suite should be measured before relying on it.

207
adapter/README.md Normal file
View File

@@ -0,0 +1,207 @@
---
base_model: Qwen/Qwen3-14B
library_name: peft
pipeline_tag: text-generation
tags:
- base_model:adapter:Qwen/Qwen3-14B
- 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 0.19.1

48
adapter/adapter_config.json Normal file
View File

@@ -0,0 +1,48 @@
{
"alora_invocation_tokens": null,
"alpha_pattern": {},
"arrow_config": null,
"auto_mapping": null,
"base_model_name_or_path": "Qwen/Qwen3-14B",
"bias": "none",
"corda_config": null,
"ensure_weight_tying": false,
"eva_config": null,
"exclude_modules": null,
"fan_in_fan_out": false,
"inference_mode": true,
"init_lora_weights": true,
"layer_replication": null,
"layers_pattern": null,
"layers_to_transform": null,
"loftq_config": {},
"lora_alpha": 64,
"lora_bias": false,
"lora_dropout": 0.0,
"lora_ga_config": null,
"megatron_config": null,
"megatron_core": "megatron.core",
"modules_to_save": null,
"peft_type": "LORA",
"peft_version": "0.19.1",
"qalora_group_size": 16,
"r": 32,
"rank_pattern": {},
"revision": null,
"target_modules": [
"gate_proj",
"v_proj",
"k_proj",
"up_proj",
"down_proj",
"o_proj",
"q_proj"
],
"target_parameters": null,
"task_type": "CAUSAL_LM",
"trainable_token_indices": null,
"use_bdlora": null,
"use_dora": false,
"use_qalora": false,
"use_rslora": false
}

3
adapter/adapter_model.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:68ee829cf0d836f834e05f7fd8048ed8728a005134279148191bcc08ac9559bb
size 513877864

89
adapter/chat_template.jinja Normal file
View File

@@ -0,0 +1,89 @@
{%- if tools %}
{{- '<|im_start|>system\n' }}
{%- if messages[0].role == 'system' %}
{{- messages[0].content + '\n\n' }}
{%- endif %}
{{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
{%- for tool in tools %}
{{- "\n" }}
{{- tool | tojson }}
{%- endfor %}
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
{%- else %}
{%- if messages[0].role == 'system' %}
{{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
{%- for message in messages[::-1] %}
{%- set index = (messages|length - 1) - loop.index0 %}
{%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
{%- set ns.multi_step_tool = false %}
{%- set ns.last_query_index = index %}
{%- endif %}
{%- endfor %}
{%- for message in messages %}
{%- if message.content is string %}
{%- set content = message.content %}
{%- else %}
{%- set content = '' %}
{%- endif %}
{%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
{%- elif message.role == "assistant" %}
{%- set reasoning_content = '' %}
{%- if message.reasoning_content is string %}
{%- set reasoning_content = message.reasoning_content %}
{%- else %}
{%- if '</think>' in content %}
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
{%- endif %}
{%- endif %}
{%- if loop.index0 > ns.last_query_index %}
{%- if loop.last or (not loop.last and reasoning_content) %}
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
{%- else %}
{{- '<|im_start|>' + message.role + '\n' + content }}
{%- endif %}
{%- else %}
{{- '<|im_start|>' + message.role + '\n' + content }}
{%- endif %}
{%- if message.tool_calls %}
{%- for tool_call in message.tool_calls %}
{%- if (loop.first and content) or (not loop.first) %}
{{- '\n' }}
{%- endif %}
{%- if tool_call.function %}
{%- set tool_call = tool_call.function %}
{%- endif %}
{{- '<tool_call>\n{"name": "' }}
{{- tool_call.name }}
{{- '", "arguments": ' }}
{%- if tool_call.arguments is string %}
{{- tool_call.arguments }}
{%- else %}
{{- tool_call.arguments | tojson }}
{%- endif %}
{{- '}\n</tool_call>' }}
{%- endfor %}
{%- endif %}
{{- '<|im_end|>\n' }}
{%- elif message.role == "tool" %}
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
{{- '<|im_start|>user' }}
{%- endif %}
{{- '\n<tool_response>\n' }}
{{- content }}
{{- '\n</tool_response>' }}
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
{{- '<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- endfor %}
{%- if add_generation_prompt %}
{{- '<|im_start|>assistant\n' }}
{%- if enable_thinking is defined and enable_thinking is false %}
{{- '<think>\n\n</think>\n\n' }}
{%- endif %}
{%- endif %}

3
adapter/tokenizer.json Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
size 11422650

29
adapter/tokenizer_config.json Normal file
View File

@@ -0,0 +1,29 @@
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": null,
"clean_up_tokenization_spaces": false,
"eos_token": "<|im_end|>",
"errors": "replace",
"extra_special_tokens": [
"<|im_start|>",
"<|im_end|>",
"<|object_ref_start|>",
"<|object_ref_end|>",
"<|box_start|>",
"<|box_end|>",
"<|quad_start|>",
"<|quad_end|>",
"<|vision_start|>",
"<|vision_end|>",
"<|vision_pad|>",
"<|image_pad|>",
"<|video_pad|>"
],
"is_local": false,
"model_max_length": 131072,
"pad_token": "<|endoftext|>",
"split_special_tokens": false,
"tokenizer_class": "Qwen2Tokenizer",
"unk_token": null
}

3
adapter/training_args.bin Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:f7b8e0e2fb5df1b2a243361e988314c091bebb39a1d6987c307d909b5412bd56
size 5201

28
adele_judge_config.json Normal file
View File

@@ -0,0 +1,28 @@
{
"inference": {
"allowed_scores": [
"1",
"2",
"3",
"4",
"5"
],
"binary_threshold": 3,
"method": "restricted_continuation_logprobs_fast"
},
"model": {
"base_model": "Qwen/Qwen3-14B",
"revision": null,
"thinking_mode": {
"apply_if_supported": true,
"enabled": false
}
},
"prompt": {
"system_prompt": "You are an answer quality evaluator. You will receive a question, a reference\nanswer, and a model response, each delimited by ### headers.\n\nRate the model response on the following scale:\n1 - Surely incorrect: the response is wrong or completely misses the point.\n2 - Likely incorrect: the response has significant errors or critical omissions.\n3 - Minimally correct / sufficient: the response is partially correct or borderline.\n4 - Likely correct: the response is mostly correct with minor issues.\n5 - Surely correct: the response is fully correct and complete.\n\nRules:\n- Base your rating strictly on correctness relative to the reference answer.\n- Ignore writing style, verbosity, or formatting unless they affect correctness.\n- If the final answer is correct but the reasoning contains substantive errors, lower the rating according to the severity of those errors.\n- Output only a single integer from 1 to 5. No explanation. No punctuation.\n"
},
"training": {
"max_seq_length": 4096,
"objective": "restricted_score_ce"
}
}

1059
adele_judge_metadata.json Normal file
View File

File diff suppressed because it is too large Load Diff

304
adele_judge_pipeline.py Normal file
View File

@@ -0,0 +1,304 @@
from __future__ import annotations
import inspect
import json
from pathlib import Path
from typing import Any
from transformers import Pipeline
THINKING_KWARG = "enable_thinking"
DEFAULT_SYSTEM_PROMPT = "Return only one score from 1 to 5. Do not explain."
DEFAULT_ALLOWED_SCORES = ["1", "2", "3", "4", "5"]
DEFAULT_BINARY_THRESHOLD = 3
def load_adele_judge_config(repo_id_or_path: str) -> dict[str, Any]:
path = Path(repo_id_or_path) / "adele_judge_config.json"
if path.exists():
return json.loads(path.read_text(encoding="utf-8"))
from huggingface_hub import hf_hub_download
downloaded = hf_hub_download(repo_id_or_path, "adele_judge_config.json")
return json.loads(Path(downloaded).read_text(encoding="utf-8"))
def load_adele_judge_config_or_default(model: Any, tokenizer: Any) -> dict[str, Any]:
candidates = [
getattr(model, "name_or_path", None),
getattr(getattr(model, "config", None), "_name_or_path", None),
getattr(tokenizer, "name_or_path", None),
getattr(tokenizer, "_name_or_path", None),
]
for candidate in candidates:
if not candidate:
continue
try:
return load_adele_judge_config(str(candidate))
except Exception:
continue
return {}
def adele_judge_settings(config: dict[str, Any] | None) -> dict[str, Any]:
config = config or {}
prompt_config = config.get("prompt", {}) if isinstance(config.get("prompt"), dict) else {}
inference_config = (
config.get("inference", {}) if isinstance(config.get("inference"), dict) else {}
)
model_config = config.get("model", {}) if isinstance(config.get("model"), dict) else {}
return {
"system_prompt": prompt_config.get("system_prompt") or DEFAULT_SYSTEM_PROMPT,
"allowed_scores": [
str(score)
for score in inference_config.get("allowed_scores", DEFAULT_ALLOWED_SCORES)
],
"binary_threshold": int(
inference_config.get("binary_threshold", DEFAULT_BINARY_THRESHOLD)
),
"thinking_mode": model_config.get("thinking_mode") or {},
}
def clean_value(value: Any, fallback: str = "N/A") -> str:
if value is None:
return fallback
text = str(value)
if not text or text.lower() == "nan":
return fallback
return text
def validate_example(inputs: Any) -> dict[str, Any]:
if not isinstance(inputs, dict):
raise ValueError("ADeLe judge input must be a mapping")
missing = []
if inputs.get("question") is None:
missing.append("question")
if inputs.get("model_response") is None:
missing.append("model_response")
reference_answer = inputs.get("reference_answer")
if reference_answer is None:
reference_answer = inputs.get("ground_truth")
if reference_answer is None:
missing.append("reference_answer or ground_truth")
if missing:
raise ValueError(f"Missing required field(s): {', '.join(missing)}")
return {
"question": inputs["question"],
"reference_answer": reference_answer,
"model_response": inputs["model_response"],
}
def build_user_message(example: dict[str, Any]) -> str:
return "\n\n".join(
[
f"### QUESTION\n{clean_value(example.get('question'))}",
f"### REFERENCE ANSWER\n{clean_value(example.get('reference_answer'))}",
f"### MODEL RESPONSE\n{clean_value(example.get('model_response'), fallback='')}",
"### SCORE\n",
]
)
def build_messages(example: dict[str, Any], system_prompt: str) -> list[dict[str, str]]:
return [
{"role": "system", "content": system_prompt.strip()},
{"role": "user", "content": build_user_message(example)},
]
def chat_template_supports_thinking(tokenizer: Any) -> bool:
apply_chat_template = getattr(tokenizer, "apply_chat_template", None)
if apply_chat_template is None:
return False
try:
signature = inspect.signature(apply_chat_template)
except (TypeError, ValueError):
return False
accepts_kwarg = any(
parameter.kind == inspect.Parameter.VAR_KEYWORD or name == THINKING_KWARG
for name, parameter in signature.parameters.items()
)
if not accepts_kwarg:
return False
template = getattr(tokenizer, "chat_template", None)
if isinstance(template, str) and THINKING_KWARG in template:
return True
candidates = [
getattr(tokenizer, "name_or_path", None),
getattr(tokenizer, "_name_or_path", None),
getattr(tokenizer, "model_name", None),
]
init_kwargs = getattr(tokenizer, "init_kwargs", None)
if isinstance(init_kwargs, dict):
candidates.extend([init_kwargs.get("name_or_path"), init_kwargs.get("tokenizer_file")])
return any("qwen3" in str(candidate).lower() for candidate in candidates if candidate)
def apply_chat_template_safe(
tokenizer: Any,
messages: list[dict[str, str]],
*,
add_generation_prompt: bool,
thinking_mode: dict[str, Any],
) -> str:
if hasattr(tokenizer, "apply_chat_template") and getattr(tokenizer, "chat_template", None):
template_kwargs = {}
enabled = thinking_mode.get("enabled")
if (
enabled is not None
and bool(thinking_mode.get("apply_if_supported", True))
and chat_template_supports_thinking(tokenizer)
):
template_kwargs[THINKING_KWARG] = bool(enabled)
return tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=add_generation_prompt,
**template_kwargs,
)
rendered = [f"<|{message['role']}|>\n{message['content']}" for message in messages]
if add_generation_prompt:
rendered.append("<|assistant|>\n")
return "\n".join(rendered)
def encode_text(tokenizer: Any, text: str) -> list[int]:
return tokenizer(text, add_special_tokens=False, truncation=False)["input_ids"]
def single_score_token_ids(tokenizer: Any, allowed_scores: list[str]) -> list[int]:
token_ids = [encode_text(tokenizer, score) for score in allowed_scores]
multi_token_scores = [
score for score, ids in zip(allowed_scores, token_ids, strict=True) if len(ids) != 1
]
if multi_token_scores:
raise ValueError(
"ADeLeJudgePipeline requires score continuations to be single tokens; "
f"multi-token scores: {multi_token_scores}"
)
return [ids[0] for ids in token_ids]
class ADeLeJudgePipeline(Pipeline):
"""HF-native custom pipeline for restricted ADeLe judge scoring."""
def __init__(
self,
*args: Any,
adele_config: dict[str, Any] | None = None,
**kwargs: Any,
) -> None:
super().__init__(*args, **kwargs)
if self.tokenizer is None:
raise ValueError("ADeLeJudgePipeline requires a tokenizer")
if getattr(self.tokenizer, "pad_token", None) is None:
self.tokenizer.pad_token = getattr(self.tokenizer, "eos_token", None)
settings = adele_judge_settings(
adele_config
if adele_config is not None
else load_adele_judge_config_or_default(self.model, self.tokenizer)
)
self.system_prompt = settings["system_prompt"]
self.allowed_scores = settings["allowed_scores"]
self.binary_threshold = settings["binary_threshold"]
self.thinking_mode = settings["thinking_mode"]
self.score_token_ids = single_score_token_ids(self.tokenizer, self.allowed_scores)
if hasattr(self.model, "eval"):
self.model.eval()
def _sanitize_parameters(
self,
**kwargs: Any,
) -> tuple[dict[str, Any], dict[str, Any], dict[str, Any]]:
return {}, {}, {}
def preprocess(self, inputs: Any) -> dict[str, Any]:
import torch
example = validate_example(inputs)
prompt = apply_chat_template_safe(
self.tokenizer,
build_messages(example, self.system_prompt),
add_generation_prompt=True,
thinking_mode=self.thinking_mode,
)
encoded = self.tokenizer(
prompt,
add_special_tokens=False,
truncation=False,
return_tensors="pt",
)
if "attention_mask" not in encoded:
encoded["attention_mask"] = torch.ones_like(encoded["input_ids"])
return {"input_ids": encoded["input_ids"], "attention_mask": encoded["attention_mask"]}
def _forward(self, model_inputs: dict[str, Any]) -> dict[str, Any]:
import torch
import torch.nn.functional as F
input_ids = model_inputs["input_ids"]
attention_mask = model_inputs["attention_mask"]
with torch.no_grad():
outputs = self.model(input_ids=input_ids, attention_mask=attention_mask)
token_positions = torch.arange(input_ids.shape[1], device=input_ids.device).unsqueeze(0)
positions = (attention_mask * token_positions).max(dim=1).values.to(dtype=torch.long)
batch_indices = torch.arange(input_ids.shape[0], device=input_ids.device)
final_logits = outputs.logits[batch_indices, positions]
score_ids = torch.tensor(self.score_token_ids, dtype=torch.long, device=final_logits.device)
score_logits = final_logits[:, score_ids]
logprobs = F.log_softmax(score_logits, dim=-1)
return {
"score_indices": torch.argmax(logprobs, dim=-1),
"probs": torch.exp(logprobs),
"logprobs": logprobs,
}
def postprocess(self, model_outputs: dict[str, Any]) -> dict[str, Any]:
import torch
score_index = int(model_outputs["score_indices"].reshape(-1)[0])
probs_tensor = model_outputs["probs"].reshape(-1, len(self.allowed_scores))[0]
logprobs_tensor = model_outputs["logprobs"].reshape(-1, len(self.allowed_scores))[0]
probs = {
score: float(prob)
for score, prob in zip(self.allowed_scores, probs_tensor.tolist(), strict=True)
}
logprobs = {
score: float(logprob)
for score, logprob in zip(self.allowed_scores, logprobs_tensor.tolist(), strict=True)
}
score = int(self.allowed_scores[score_index])
sorted_logprobs = torch.sort(logprobs_tensor).values
margin = (
float(sorted_logprobs[-1] - sorted_logprobs[-2])
if len(sorted_logprobs) > 1
else 0.0
)
entropy = float(
-(probs_tensor * torch.log(torch.clamp(probs_tensor, min=1e-12))).sum()
)
return {
"score": score,
"label": "CORRECT" if score >= self.binary_threshold else "INCORRECT",
"probs": probs,
"logprobs": logprobs,
"confidence": max(probs.values()),
"margin": margin,
"entropy": entropy,
}

89
chat_template.jinja Normal file
View File

@@ -0,0 +1,89 @@
{%- if tools %}
{{- '<|im_start|>system\n' }}
{%- if messages[0].role == 'system' %}
{{- messages[0].content + '\n\n' }}
{%- endif %}
{{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
{%- for tool in tools %}
{{- "\n" }}
{{- tool | tojson }}
{%- endfor %}
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
{%- else %}
{%- if messages[0].role == 'system' %}
{{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
{%- for message in messages[::-1] %}
{%- set index = (messages|length - 1) - loop.index0 %}
{%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
{%- set ns.multi_step_tool = false %}
{%- set ns.last_query_index = index %}
{%- endif %}
{%- endfor %}
{%- for message in messages %}
{%- if message.content is string %}
{%- set content = message.content %}
{%- else %}
{%- set content = '' %}
{%- endif %}
{%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
{%- elif message.role == "assistant" %}
{%- set reasoning_content = '' %}
{%- if message.reasoning_content is string %}
{%- set reasoning_content = message.reasoning_content %}
{%- else %}
{%- if '</think>' in content %}
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
{%- endif %}
{%- endif %}
{%- if loop.index0 > ns.last_query_index %}
{%- if loop.last or (not loop.last and reasoning_content) %}
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
{%- else %}
{{- '<|im_start|>' + message.role + '\n' + content }}
{%- endif %}
{%- else %}
{{- '<|im_start|>' + message.role + '\n' + content }}
{%- endif %}
{%- if message.tool_calls %}
{%- for tool_call in message.tool_calls %}
{%- if (loop.first and content) or (not loop.first) %}
{{- '\n' }}
{%- endif %}
{%- if tool_call.function %}
{%- set tool_call = tool_call.function %}
{%- endif %}
{{- '<tool_call>\n{"name": "' }}
{{- tool_call.name }}
{{- '", "arguments": ' }}
{%- if tool_call.arguments is string %}
{{- tool_call.arguments }}
{%- else %}
{{- tool_call.arguments | tojson }}
{%- endif %}
{{- '}\n</tool_call>' }}
{%- endfor %}
{%- endif %}
{{- '<|im_end|>\n' }}
{%- elif message.role == "tool" %}
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
{{- '<|im_start|>user' }}
{%- endif %}
{{- '\n<tool_response>\n' }}
{{- content }}
{{- '\n</tool_response>' }}
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
{{- '<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- endfor %}
{%- if add_generation_prompt %}
{{- '<|im_start|>assistant\n' }}
{%- if enable_thinking is defined and enable_thinking is false %}
{{- '<think>\n\n</think>\n\n' }}
{%- endif %}
{%- endif %}

85
config.json Normal file
View File

@@ -0,0 +1,85 @@
{
"architectures": [
"Qwen3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 151643,
"custom_pipelines": {
"adele-judge": {
"impl": "adele_judge_pipeline.ADeLeJudgePipeline",
"pt": [
"AutoModelForCausalLM"
],
"tf": [],
"type": "text"
}
},
"dtype": "bfloat16",
"eos_token_id": 151645,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17408,
"layer_types": [
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention"
],
"max_position_embeddings": 40960,
"max_window_layers": 40,
"model_type": "qwen3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 8,
"pad_token_id": null,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 1000000,
"rope_type": "default"
},
"sliding_window": null,
"tie_word_embeddings": false,
"transformers_version": "5.5.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151936
}

5
generation_config.json Normal file
View File

@@ -0,0 +1,5 @@
{
"do_sample": false,
"max_new_tokens": 1,
"num_beams": 1
}

3
model-00001-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:5739c7705041195574706abc2af5e12c66c2281e1a93dd1419dd2439c0fd1a41
size 4978180848

3
model-00002-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:418628dcaf69131e39dff0aa22339698bf3111fd402a0c677ff8b1fb08a146de
size 4917988408

3
model-00003-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:7e63ccd5fd8e91297f66be63117190fbe6c5ad636c669d33263f490c42270e3a
size 4991388720

3
model-00004-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:483606985498d68d803f7761870f76b0df2965034149f2d4af7f7c93bf707c3b
size 4917988496

3
model-00005-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:acd0b2c97fb9d8fda9360f6293ccf899914f43d54ab8bcba3557ac62412451bd
size 4991388720

3
model-00006-of-00006.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:2e2e7fd55cd67b4370f657b6f1d3ee9673f8a8657e4a90eaf873bed6661744ad
size 4739730440

451
model.safetensors.index.json Normal file
View File

@@ -0,0 +1,451 @@
{
"metadata": {
"total_parameters": 14768307200,
"total_size": 29536614400
},
"weight_map": {
"lm_head.weight": "model-00001-of-00006.safetensors",
"model.embed_tokens.weight": "model-00001-of-00006.safetensors",
"model.layers.0.input_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.0.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.k_norm.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.q_norm.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.0.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.input_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.1.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.k_norm.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.q_norm.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.1.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.10.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.10.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.10.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.10.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.10.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.10.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.11.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.11.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.12.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.12.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.13.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.13.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.14.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.14.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.15.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.15.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.16.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.q_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.16.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.17.input_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.17.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.17.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.17.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.17.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
"model.layers.17.self_attn.k_norm.weight": "model-00003-of-00006.safetensors",
"model.layers.17.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
"model.layers.17.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.17.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.17.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.17.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.18.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.18.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.19.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.19.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.2.input_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.2.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.2.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.2.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.2.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
"model.layers.2.self_attn.k_norm.weight": "model-00001-of-00006.safetensors",
"model.layers.2.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
"model.layers.2.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.2.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.2.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.2.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.20.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.20.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.20.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.21.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.21.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.22.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.22.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.23.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.23.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.24.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.k_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.q_norm.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.24.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.25.input_layernorm.weight": "model-00004-of-00006.safetensors",
"model.layers.25.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
"model.layers.25.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.25.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.25.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.25.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.26.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.26.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.27.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.27.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.28.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.28.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.29.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.29.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.3.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.3.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.3.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.30.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.30.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.30.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.31.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.q_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.31.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.32.input_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.32.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.32.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.32.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.32.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
"model.layers.32.self_attn.k_norm.weight": "model-00005-of-00006.safetensors",
"model.layers.32.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
"model.layers.32.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.32.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.32.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.32.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.33.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.33.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.34.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.34.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.35.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.35.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.36.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.36.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.37.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.37.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.38.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.38.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.input_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.39.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.k_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.q_norm.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.39.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
"model.layers.4.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.4.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.4.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.5.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.5.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.6.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.6.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.7.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.7.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.8.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.8.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.input_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.9.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.k_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.q_norm.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
"model.layers.9.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
"model.norm.weight": "model-00006-of-00006.safetensors"
}
}

3
tokenizer.json Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
size 11422650

29
tokenizer_config.json Normal file
View File

@@ -0,0 +1,29 @@
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": null,
"clean_up_tokenization_spaces": false,
"eos_token": "<|im_end|>",
"errors": "replace",
"extra_special_tokens": [
"<|im_start|>",
"<|im_end|>",
"<|object_ref_start|>",
"<|object_ref_end|>",
"<|box_start|>",
"<|box_end|>",
"<|quad_start|>",
"<|quad_end|>",
"<|vision_start|>",
"<|vision_end|>",
"<|vision_pad|>",
"<|image_pad|>",
"<|video_pad|>"
],
"is_local": true,
"model_max_length": 131072,
"pad_token": "<|endoftext|>",
"split_special_tokens": false,
"tokenizer_class": "Qwen2Tokenizer",
"unk_token": null
}

168
training_config.yaml Normal file
View File

@@ -0,0 +1,168 @@
project:
run_name: qwen3_14b_restricted_score_ce
output_dir: runs/qwen3_14b_restricted_score_ce
seed: 42
data:
path: data/processed/response_scores.parquet
prepared_dir: null
columns:
question: question
reference_answer: ground_truth
response: response
judge_1_score: score_gpt4o
judge_2_score: score_sonnet
model_id: model_id
benchmark: benchmark
task: task
example_id: instance_id
source: source
filters:
max_disagreement: 1
max_response_tokens: 4096
on_sequence_overflow: skip
preprocessing_num_workers: 40
tokenizers_parallelism: true
token_length_batch_size: 2048
model:
model_name_or_path: Qwen/Qwen3-14B
revision: null
attn_implementation: sdpa
adapter_path: null
trust_remote_code: true
thinking_mode:
enabled: false
apply_if_supported: true
prompt:
system_prompt: 'You are an answer quality evaluator. You will receive a question,
a reference
answer, and a model response, each delimited by ### headers.
Rate the model response on the following scale:
1 - Surely incorrect: the response is wrong or completely misses the point.
2 - Likely incorrect: the response has significant errors or critical omissions.
3 - Minimally correct / sufficient: the response is partially correct or borderline.
4 - Likely correct: the response is mostly correct with minor issues.
5 - Surely correct: the response is fully correct and complete.
Rules:
- Base your rating strictly on correctness relative to the reference answer.
- Ignore writing style, verbosity, or formatting unless they affect correctness.
- If the final answer is correct but the reasoning contains substantive errors,
lower the rating according to the severity of those errors.
- Output only a single integer from 1 to 5. No explanation. No punctuation.
'
split:
mode: fixed_by_model
validation_models:
- gemini-3-flash
- DK-R1-Dist-Qwen-14B
- llama3d2-3b
train_models: auto_except_val_test
held_out_model: null
lomo_validation_fraction: 0.05
lomo_validation_max_examples: 30000
lomo_validation_seed: 42
training:
max_seq_length: 4096
load_in_4bit: true
dtype: bfloat16
objective: restricted_score_ce
loss:
type: ce_5way
lambda_binary: 0.5
class_weights: null
class_weighting: null
score_class_weights: null
lora_r: 32
lora_alpha: 64
lora_dropout: 0.0
target_modules: auto
learning_rate: 3.0e-05
num_train_epochs: 1
per_device_train_batch_size: 2
per_device_eval_batch_size: 2
gradient_accumulation_steps: 4
warmup_ratio: 0.03
lr_scheduler_type: cosine
weight_decay: 0.0
optim: adamw_8bit
packing: false
cache_tokenized_datasets: true
eval_subset_size: null
eval_subset_strategy: stratified
eval_subset_stratify_columns:
- model_id
- target_score
train_sampling_strategy: random
length_column_name: length
logging_steps: 10
eval_steps: 500
save_steps: 500
save_total_limit: 10
seed: 42
resume_from_checkpoint: null
eval_subset_seed: 42
max_grad_norm: 1.0
distributed:
enabled: true
strategy: ddp
backend: nccl
mixed_precision: bf16
gradient_checkpointing: false
find_unused_parameters: false
fsdp:
sharding_strategy: full_shard
transformer_layer_cls_to_wrap: null
activation_checkpointing: true
use_orig_params: true
deepspeed:
zero_stage: 2
offload_optimizer_device: none
offload_param_device: none
stage3_gather_16bit_weights_on_model_save: true
gradient_clipping: auto
config_overrides: {}
inference:
allowed_scores:
- '1'
- '2'
- '3'
- '4'
- '5'
binary_threshold: 3
method: restricted_continuation_logprobs_fast
generation_fallback: false
batch_size: 64
require_adapter: false
allow_base_model: true
evaluation:
length_buckets:
- 0
- 256
- 512
- 1024
- 2048
- 3072
- 4096
- 1000000000
hub:
repo_id: null
private: false
commit_message: Upload ADeLe distilled judge
local_checkpoint_dir: null
output_staging_dir: null
create_pr: false
max_shard_size: 5GB