初始化项目,由ModelHub XC社区提供模型

Model: NotoriousH2/Qwen3-4B-Calendar-Agent-RLVR
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-10-15 03:26:19 +08:00
commit d8f7269e8e
16 changed files with 914 additions and 0 deletions

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

140
README.md Normal file
View File

@@ -0,0 +1,140 @@
---
license: apache-2.0
language:
- ko
library_name: transformers
pipeline_tag: text-generation
base_model: NotoriousH2/Qwen3-4B-Calendar-Agent-SFT
datasets:
- NotoriousH2/calendar-agent-benchmark
tags:
- qwen3
- tool-calling
- function-calling
- agents
- grpo
- rlvr
- reinforcement-learning
- calendar
- korean
---
# Qwen3-4B Calendar Agent RLVR
Qwen3-4B Calendar Agent RLVR는 한국어 일정 관리 도구를 호출하는 4B 파라미터 병합 모델입니다.
Calendar 도구의 실행 궤적과 최종 상태를 보상으로 평가하여 GRPO로 학습했습니다.
다음 행동을 학습 범위에 포함합니다.
- 일정 조회, 생성, 수정, 삭제
- 필수 정보가 없을 때 사용자에게 정보 요청
- 삭제 전 사용자 승인 확인
- 일정 충돌 확인과 빈 시간 생성
- 여러 참석자의 공통 가능 시간 탐색
## 1. 사용
저장소 루트에는 병합 모델이 있으며, `adapter/`에는 LoRA 어댑터가 있습니다.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "NotoriousH2/Qwen3-4B-Calendar-Agent-RLVR"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id,
dtype="auto",
device_map="auto",
)
tools = [
{
"type": "function",
"function": {
"name": "search_events",
"description": "지정한 날짜의 일정을 조회합니다.",
"parameters": {
"type": "object",
"properties": {
"date": {"type": "string", "description": "YYYY-MM-DD"}
},
"required": ["date"],
},
},
}
]
messages = [
{
"role": "system",
"content": "사용자 요청을 처리하는 데 필요한 일정 도구를 호출하세요.",
},
{"role": "user", "content": "2039-07-05 일정을 알려줘."},
]
prompt = tokenizer.apply_chat_template(
messages,
tools=tools,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False,
)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512)
response = outputs[0, inputs.input_ids.shape[1] :]
print(tokenizer.decode(response, skip_special_tokens=False))
```
모델은 도구 호출 정보를 생성합니다.
호출 실행과 도구 결과 전달은 호스트 애플리케이션에서 처리해야 합니다.
## 2. 학습 보상
같은 요청에서 16개 응답을 생성하고, 그룹 안의 보상 차이로 Advantage를 계산했습니다.
| 보상 | 가중치 | 평가 대상 |
|---|---:|---|
| 안전 규칙 | 0.25 | 필수 정보와 사용자 승인 준수 |
| 절차 진행 | 0.35 | 조회, 확인, 성공한 도구 실행 |
| 최종 성공 | 0.40 | 최종 Calendar 상태와 사용자 요청의 일치 |
전체 행동군을 다룬 첫 학습 단계에서 96개 그룹 중 68개 그룹이 서로 다른 보상을 포함했습니다.
최종 성공 여부가 같은 응답도 절차 진행 점수로 구분하여 학습 신호를 구성했습니다.
## 3. 평가
400개 합성 시나리오에서 실제 도구 루프를 실행하고 최종 상태를 평가했습니다.
| 모델 | 유효 도구 호출 | 정책 준수 | Task Success | 평균 도구 호출 |
|---|---:|---:|---:|---:|
| Qwen3-4B Base | 100.00% | 80.00% | 20.00% | 0.80 |
| Calendar Agent SFT | 100.00% | 89.25% | 89.00% | 2.55 |
| Calendar Agent RLVR | 100.00% | 99.25% | 99.00% | 2.38 |
SFT와 사례별로 비교하면 RLVR 모델은 41개 실패를 성공으로 전환했고, 1개 성공을 실패로 전환했습니다.
순증은 40개이며 전체 Task Success 차이는 10.00%p입니다.
템플릿별 Task Success는 다음과 같습니다.
| 템플릿 | SFT | RLVR | 전체 |
|---|---:|---:|---:|
| 필수 정보 요청 | 60 | 80 | 80 |
| 삭제 승인 요청 | 59 | 80 | 80 |
| 충돌 후 생성 | 77 | 76 | 80 |
| 승인 후 삭제 | 80 | 80 | 80 |
| 다중 참석자 생성 | 80 | 80 | 80 |
집계값은 [evaluation.json](./evaluation.json)에 있습니다.
## 4. 행동 사례
필수 정보가 없는 생성 요청에서는 임의의 시각을 정해 일정을 생성하지 않고 시작 시각과 소요 시간을 요청합니다.
충돌 시 대체 시간 등록을 승인한 요청에서는 기존 일정을 조회하고 빈 시간을 찾은 뒤, 충돌하지 않는 시간에 일정을 생성합니다.
## 5. 제한
이 모델은 `Asia/Seoul` 시간대의 합성 시나리오와 메모리 Calendar 환경에서 학습하고 평가했습니다.
실제 일정 서비스의 인증, 권한, 개인정보 처리, 네트워크 장애는 평가 범위에 포함하지 않습니다.
충돌 후 생성은 SFT 77/80, RLVR 76/80으로 다른 템플릿보다 낮습니다.
외부 Calendar 시스템에 적용할 때는 충돌 복구와 도구 오류 이후의 재시도를 별도로 평가해야 합니다.

209
adapter/README.md Normal file
View File

@@ -0,0 +1,209 @@
---
base_model: artifacts/calendar_sft/merged
library_name: peft
pipeline_tag: text-generation
tags:
- base_model:adapter:artifacts/calendar_sft/merged
- grpo
- lora
- transformers
- trl
---
# 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 0.20.0

View File

@@ -0,0 +1,50 @@
{
"alora_invocation_tokens": null,
"alpha_pattern": {},
"arrow_config": null,
"auto_mapping": null,
"base_model_name_or_path": "artifacts/calendar_sft/merged",
"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": 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": "0.20.0",
"qalora_group_size": 16,
"r": 16,
"rank_pattern": {},
"revision": null,
"target_modules": [
"q_proj",
"down_proj",
"gate_proj",
"o_proj",
"v_proj",
"k_proj",
"up_proj"
],
"target_parameters": null,
"task_type": "CAUSAL_LM",
"trainable_token_indices": null,
"use_bdlora": null,
"use_dora": false,
"use_qalora": false,
"use_rslora": false,
"velora_config": null
}

View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:0a63322168467878b8e7b11334d96051085539e6a85793653257625ba4618c95
size 132187888

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

View File

@@ -0,0 +1,57 @@
{
"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,
"local_files_only": false,
"model_max_length": 131072,
"pad_token": "<|endoftext|>",
"response_template": {
"defaults": {
"role": "assistant"
},
"fields": {
"content": {
"close_pattern": "<\\|im_end\\|>\\s*",
"content": "text"
},
"reasoning_content": {
"close_pattern": "</think>\\s*",
"content": "text",
"open": "<think>"
},
"tool_calls": {
"close_pattern": "</tool_call>\\s*",
"content": "json",
"open": "<tool_call>",
"repeats": true,
"transform": {
"function": "{content}",
"type": "function"
}
}
},
"start_anchor": "<|im_start|>assistant\n"
},
"split_special_tokens": false,
"tokenizer_class": "Qwen2Tokenizer",
"unk_token": null
}

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 %}

71
config.json Normal file
View File

@@ -0,0 +1,71 @@
{
"architectures": [
"Qwen3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": null,
"dtype": "bfloat16",
"eos_token_id": 151645,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 2560,
"initializer_range": 0.02,
"intermediate_size": 9728,
"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"
],
"max_position_embeddings": 40960,
"max_window_layers": 36,
"model_type": "qwen3",
"num_attention_heads": 32,
"num_hidden_layers": 36,
"num_key_value_heads": 8,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 1000000,
"rope_type": "default"
},
"sliding_window": null,
"tie_word_embeddings": true,
"transformers_version": "5.14.1",
"use_cache": false,
"use_sliding_window": false,
"vocab_size": 151936
}

92
evaluation.json Normal file
View File

@@ -0,0 +1,92 @@
{
"task": "calendar_tool_calling",
"timezone": "Asia/Seoul",
"samples": 400,
"models": [
{
"model": "Qwen3-4B Base",
"valid_tool_call": 1.0,
"policy_compliance": 0.8,
"task_success": 0.2,
"average_tool_calls": 0.8
},
{
"model": "Calendar Agent SFT",
"valid_tool_call": 1.0,
"policy_compliance": 0.8925,
"task_success": 0.89,
"average_tool_calls": 2.5525
},
{
"model": "Calendar Agent RLVR",
"valid_tool_call": 1.0,
"policy_compliance": 0.9925,
"task_success": 0.99,
"average_tool_calls": 2.3775
}
],
"paired_transition": {
"task_success": {
"gains": 41,
"losses": 1,
"net_gain": 40
},
"policy_compliance": {
"gains": 41,
"losses": 1,
"net_gain": 40
}
},
"templates": [
{
"template": "missing_time_clarification",
"samples": 80,
"sft_successes": 60.0,
"rlvr_successes": 80.0
},
{
"template": "delete_confirmation",
"samples": 80,
"sft_successes": 59.0,
"rlvr_successes": 80.0
},
{
"template": "conflict_then_create",
"samples": 80,
"sft_successes": 77.0,
"rlvr_successes": 76.0
},
{
"template": "confirmed_delete",
"samples": 80,
"sft_successes": 80.0,
"rlvr_successes": 80.0
},
{
"template": "multi_attendee_create",
"samples": 80,
"sft_successes": 80.0,
"rlvr_successes": 80.0
}
],
"training_signal": {
"rollouts_per_group": 16,
"groups": 96,
"active_groups": 68,
"zero_variance_groups": 28,
"reward_components": [
{
"name": "safety",
"weight": 0.25
},
{
"name": "policy_progress",
"weight": 0.35
},
{
"name": "semantic_final_success",
"weight": 0.4
}
]
}
}

12
generation_config.json Normal file
View File

@@ -0,0 +1,12 @@
{
"do_sample": true,
"eos_token_id": [
151645,
151643
],
"pad_token_id": 151643,
"temperature": 0.6,
"top_k": 20,
"top_p": 0.95,
"transformers_version": "5.14.1"
}

26
metrics.json Normal file
View File

@@ -0,0 +1,26 @@
{
"task": "calendar",
"models": [
{
"model": "Qwen3-4B Base",
"valid_tool_call": 1.0,
"policy_compliance": 0.8,
"task_success": 0.2,
"average_tool_calls": 0.8
},
{
"model": "Calendar Agent SFT",
"valid_tool_call": 1.0,
"policy_compliance": 0.8925,
"task_success": 0.89,
"average_tool_calls": 2.5525
},
{
"model": "Calendar Agent RLVR",
"valid_tool_call": 1.0,
"policy_compliance": 0.9925,
"task_success": 0.99,
"average_tool_calls": 2.3775
}
]
}

3
model.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:1c4bab462e4538b946e6b17239a51c535fbc16e6f1174eb80afd139d992ad41c
size 8044982080

3
tokenizer.json Normal file
View File

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

30
tokenizer_config.json Normal file
View File

@@ -0,0 +1,30 @@
{
"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,
"local_files_only": true,
"model_max_length": 131072,
"pad_token": "<|endoftext|>",
"split_special_tokens": false,
"tokenizer_class": "Qwen2Tokenizer",
"unk_token": null
}