adaamko commited on
Commit
51b4c8b
·
verified ·
1 Parent(s): a8a8e5e

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,290 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ language:
4
+ - en
5
+ base_model: Qwen/Qwen3.5-2B
6
+ tags:
7
+ - tool-output-pruning
8
+ - context-engineering
9
+ - context-pruning
10
+ - code-agent
11
+ - squeez
12
+ - qwen3.5
13
+ pipeline_tag: text-generation
14
+ library_name: transformers
15
+ datasets:
16
+ - KRLabsOrg/tool-output-extraction-swebench
17
+ ---
18
+
19
+ <p align="center">
20
+ <img src="https://github.com/KRLabsOrg/squeez/blob/main/assets/squeez_mascot.png?raw=true" alt="Squeez" width="250"/>
21
+ <br><em>Squeeze out the juice, leave the pulp behind.</em>
22
+ </p>
23
+
24
+ # Squeez-2B
25
+
26
+ LLM coding agents spend **80-95% of their context window** on irrelevant tool output — passing test names, boilerplate headers, unchanged files. Squeez reads the raw output alongside a task description and returns **only the lines the agent needs to read next**, compressing tool output by ~91% on average while keeping 86% of the relevant information.
27
+
28
+ Unlike keyword search (BM25) or generic semantic highlighting, Squeez is trained specifically on tool output from real software engineering workflows — test logs, grep results, build errors, git diffs, stack traces, and more.
29
+
30
+ ## What is Squeez?
31
+
32
+ Squeez is a **tool output pruner for coding agents**. When an agent runs a tool (pytest, grep, git log, npm build, kubectl, etc.), the output is often hundreds of lines — but only a handful matter for the current task. Squeez acts as a filter between the tool and the agent's context window:
33
+
34
+ ```
35
+ Tool output (500 lines) → Squeez → Relevant lines (30 lines) → Agent context
36
+ ```
37
+
38
+ This model (Squeez-2B) is a generative approach: [Qwen 3.5 2B](https://huggingface.co/Qwen/Qwen3.5-2B) fine-tuned to extract verbatim relevant lines from tool output, given a task-specific query.
39
+
40
+ ### Why a small fine-tuned model?
41
+
42
+ - **Fast**: 2B parameters — runs on a single GPU or even CPU, serves via vLLM at high throughput
43
+ - **Accurate**: Outperforms a 35B MoE model (Qwen 3.5 35B A3B) at zero-shot by **+13% Span F1**
44
+ - **Faithful**: Returns verbatim lines only — no rewriting, no hallucination, no summarization
45
+ - **Drop-in**: Works as a CLI pipe, Python library, or vLLM server — integrates with any agent framework
46
+
47
+ ## Evaluation
48
+
49
+ Evaluated on 617 held-out test samples from SWE-bench repositories, across 14 tool types:
50
+
51
+ ### Squeez-2B vs. generative models
52
+
53
+ | Model | Span P | Span R | Span F1 | Exact Match | Fuzzy F1 | Partial Overlap | Empty Acc | ROUGE-L | Compression |
54
+ |-------|--------|--------|---------|-------------|----------|-----------------|-----------|---------|-------------|
55
+ | **Squeez-2B** | **0.8043** | **0.8624** | **0.7895** | **0.4911** | **0.8035** | **0.9189** | **0.9676** | **0.7208** | 0.9150 |
56
+ | Qwen 3.5 35B A3B (zero-shot) | 0.7402 | 0.7498 | 0.7000 | 0.3922 | 0.7254 | 0.8347 | 0.9157 | 0.7151 | 0.9177 |
57
+ | Qwen 3.5 2B (untrained) | 0.4154 | 0.5299 | 0.4075 | 0.1945 | 0.5482 | 0.7683 | 0.9157 | 0.5481 | 0.8197 |
58
+
59
+ ### Squeez-2B vs. naive baselines
60
+
61
+ | Model | Span P | Span R | Span F1 | Exact Match | Fuzzy F1 | Partial Overlap | Empty Acc | ROUGE-L | Compression |
62
+ |-------|--------|--------|---------|-------------|----------|-----------------|-----------|---------|-------------|
63
+ | **Squeez-2B** | **0.8043** | **0.8624** | **0.7895** | **0.4911** | **0.8035** | **0.9189** | **0.9676** | **0.7208** | 0.9150 |
64
+ | BM25 (10%) | 0.1277 | 0.2172 | 0.1314 | 0.0146 | 0.2314 | 0.5883 | 0.8981 | 0.2073 | 0.9036 |
65
+ | First-N (10%) | 0.0741 | 0.1445 | 0.0798 | 0.0194 | 0.1570 | 0.4389 | 0.9175 | 0.1370 | 0.9055 |
66
+ | Random (10%) | 0.0738 | 0.1009 | 0.0697 | 0.0113 | 0.1966 | 0.4984 | 0.9061 | 0.1397 | 0.9067 |
67
+ | Last-N (10%) | 0.0496 | 0.0503 | 0.0407 | 0.0129 | 0.1393 | 0.3916 | 0.8560 | 0.1173 | 0.9130 |
68
+
69
+ ### Metric definitions
70
+
71
+ - **Span F1**: strict line-level set overlap between predicted and gold relevant lines
72
+ - **Fuzzy F1**: same as Span F1 but with fuzzy substring matching (threshold 0.5)
73
+ - **Partial Overlap**: fraction of gold lines that have any overlap with predictions
74
+ - **Empty Accuracy**: correctly predicting empty vs non-empty output (tool returned nothing relevant)
75
+ - **Compression**: fraction of input removed (higher = more aggressive pruning)
76
+
77
+ ## Quick Start
78
+
79
+ ### With vLLM (recommended)
80
+
81
+ ```bash
82
+ # Start the server
83
+ pip install vllm
84
+ vllm serve KRLabsOrg/squeez-2b --dtype bfloat16 --max-model-len 16384
85
+
86
+ # Use from squeez CLI
87
+ pip install squeez
88
+ export SQUEEZ_SERVER_URL=http://localhost:8000/v1
89
+ cat output.txt | squeez "find the bug"
90
+
91
+ # Or pipe directly
92
+ python -m pytest tests/ -v 2>&1 | squeez "find the test failure related to authentication"
93
+ ```
94
+
95
+ vLLM gives you batched inference, continuous batching, and high throughput — ideal when multiple agents or tools are running concurrently.
96
+
97
+ ### With squeez (local, no server)
98
+
99
+ ```bash
100
+ pip install squeez
101
+
102
+ # Downloads and runs the model locally (no GPU server needed)
103
+ squeez "Find the failing traceback block" --input-file output.txt
104
+ ```
105
+
106
+ > **Note:** Local mode loads the model on every call. Fine for one-off use, but for repeated calls (e.g. an agent piping every tool through squeez), use vLLM — the model stays warm in memory.
107
+
108
+ ### With transformers
109
+
110
+ ```python
111
+ from transformers import AutoModelForCausalLM, AutoTokenizer
112
+ import torch
113
+
114
+ model_name = "KRLabsOrg/squeez-2b"
115
+ tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
116
+ model = AutoModelForCausalLM.from_pretrained(
117
+ model_name,
118
+ torch_dtype=torch.bfloat16,
119
+ device_map="auto",
120
+ trust_remote_code=True,
121
+ )
122
+
123
+ messages = [
124
+ {"role": "system", "content": (
125
+ "You prune verbose tool output for a coding agent. "
126
+ "Given a focused extraction query and one tool output, return only the "
127
+ "smallest verbatim evidence block(s) the agent should read next. "
128
+ "Return the kept text inside <relevant_lines> tags. "
129
+ "Do not rewrite, summarize, or invent lines."
130
+ )},
131
+ {"role": "user", "content": (
132
+ "<query>\nFix the failing authentication test\n</query>\n"
133
+ "<tool_output>\n"
134
+ "PASSED tests/test_login.py::test_valid_credentials\n"
135
+ "FAILED tests/test_login.py::test_token_refresh - AssertionError: expected 200 got 401\n"
136
+ "PASSED tests/test_login.py::test_logout\n"
137
+ "PASSED tests/test_login.py::test_rate_limiting\n"
138
+ "\n</tool_output>"
139
+ )},
140
+ ]
141
+
142
+ prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
143
+ inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
144
+
145
+ with torch.no_grad():
146
+ outputs = model.generate(**inputs, max_new_tokens=512, temperature=0.1, do_sample=True)
147
+
148
+ response = tokenizer.decode(outputs[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True)
149
+ print(response)
150
+ ```
151
+
152
+ **Output:**
153
+ ```xml
154
+ <relevant_lines>
155
+ FAILED tests/test_login.py::test_token_refresh - AssertionError: expected 200 got 401
156
+ </relevant_lines>
157
+ ```
158
+
159
+ ### Python API (with squeez)
160
+
161
+ ```python
162
+ from squeez.inference.extractor import ToolOutputExtractor
163
+
164
+ # Loads this model locally
165
+ extractor = ToolOutputExtractor(model_path="KRLabsOrg/squeez-2b")
166
+
167
+ # Or connect to a vLLM server
168
+ extractor = ToolOutputExtractor(base_url="http://localhost:8000/v1")
169
+
170
+ filtered = extractor.extract(
171
+ task="Find the referer validation block",
172
+ tool_output=raw_output,
173
+ )
174
+ print(filtered)
175
+ ```
176
+
177
+ ## Input / Output Format
178
+
179
+ **Input** — chat format with system prompt:
180
+
181
+ ```
182
+ System: You prune verbose tool output for a coding agent. Given a focused
183
+ extraction query and one tool output, return only the smallest verbatim
184
+ evidence block(s) the agent should read next. Return the kept text inside
185
+ <relevant_lines> tags. Do not rewrite, summarize, or invent lines.
186
+
187
+ User: <query>{task_description}</query>
188
+ <tool_output>{raw_tool_output}</tool_output>
189
+ ```
190
+
191
+ **Output** — verbatim relevant lines wrapped in XML:
192
+
193
+ ```xml
194
+ <relevant_lines>
195
+ {only the lines that matter, copied verbatim}
196
+ </relevant_lines>
197
+ ```
198
+
199
+ If no lines are relevant, the model returns empty tags: `<relevant_lines>\n</relevant_lines>`.
200
+
201
+ ## Supported Tool Types
202
+
203
+ The model was trained on 14 tool types from SWE-bench repositories:
204
+
205
+ | Tool type | Description | Example |
206
+ |-----------|-------------|---------|
207
+ | `test_output` | pytest / unittest output | Test failures, tracebacks, assertion errors |
208
+ | `read_file` | File contents | Source code, config files |
209
+ | `grep` | Search results | Pattern matches across files |
210
+ | `git_diff` | Code changes | Diffs between commits or branches |
211
+ | `git_log` | Commit history | Relevant commits |
212
+ | `git_blame` | Line-level attribution | Who changed what |
213
+ | `ls` | Directory listings | File structure |
214
+ | `python` | Python REPL output | Script output, errors |
215
+ | `curl` | HTTP responses | API responses, documentation |
216
+ | `build_output` | Build logs | Compilation errors, warnings |
217
+ | `lint_output` | Linter output | Style/type violations |
218
+ | `pip_install` | Package manager output | Dependency errors |
219
+ | `type_check` | Type checker output | mypy/pyright errors |
220
+ | `coverage` | Coverage reports | Uncovered lines |
221
+
222
+ ## Training Details
223
+
224
+ | Parameter | Value |
225
+ |-----------|-------|
226
+ | Base model | [Qwen/Qwen3.5-2B](https://huggingface.co/Qwen/Qwen3.5-2B) |
227
+ | Fine-tuning method | LoRA (r=16, alpha=32) via [Unsloth](https://github.com/unslothai/unsloth) |
228
+ | Training data | Squeez v3 — 10,508 samples from [SWE-bench](https://swe-bench.github.io/) |
229
+ | Epochs | 3 (best checkpoint at epoch 1.5) |
230
+ | Max sequence length | 16,384 tokens |
231
+ | Learning rate | 2e-4 |
232
+ | Batch size | 8 (effective 32 with 4x gradient accumulation) |
233
+ | Warmup | 5% of steps |
234
+ | Weight decay | 0.01 |
235
+ | Checkpoint selection | Best validation Span F1 |
236
+
237
+ ### Data generation
238
+
239
+ Training data was generated by running 14 types of tool calls on SWE-bench repositories and using a teacher model to label the relevant lines. Each sample contains:
240
+ - A focused extraction query (what the agent needs to find)
241
+ - Raw tool output (as the agent would see it)
242
+ - Gold relevant lines (the minimal set the agent should read)
243
+
244
+ Dataset: [KRLabsOrg/tool-output-extraction-swebench](https://huggingface.co/datasets/KRLabsOrg/tool-output-extraction-swebench)
245
+
246
+ ## Limitations
247
+
248
+ - Trained primarily on Python/SWE-bench data — works best on software engineering tool output, though the prompt format generalizes to other domains
249
+ - Not designed for general-purpose text summarization or question answering
250
+ - Very short outputs (<5 lines) may be returned unchanged
251
+ - Max input length is 16,384 tokens — longer outputs should be chunked
252
+
253
+ ## Use with coding agents
254
+
255
+ Add to your agent's system instructions (e.g. `CLAUDE.md` for Claude Code):
256
+
257
+ ```
258
+ Always pipe shell commands through squeez and tell exactly what you want to know.
259
+
260
+ Examples:
261
+ - `bun test 2>&1 | squeez "did the tests pass?"`
262
+ - `git log --oneline -50 | squeez "find the commit that broke CSRF"`
263
+ - `cat src/auth/middleware.py | squeez "find the referer validation logic"`
264
+
265
+ Do NOT use squeez when:
266
+ - You need exact, uncompressed output (e.g. writing a patch)
267
+ - The command is interactive
268
+ ```
269
+
270
+ ## Citation
271
+
272
+ ```bibtex
273
+ @software{kovacs2026squeez,
274
+ title={Squeez: Compressing Tool Output for LLM Coding Agents},
275
+ author={Adam Kovacs},
276
+ year={2026},
277
+ url={https://github.com/KRLabsOrg/squeez}
278
+ }
279
+ ```
280
+
281
+ ## License
282
+
283
+ Apache 2.0
284
+
285
+ ## Acknowledgments
286
+
287
+ - [Qwen](https://huggingface.co/Qwen) for the Qwen 3.5 2B base model
288
+ - [Unsloth](https://github.com/unslothai/unsloth) for efficient LoRA training
289
+ - [SWE-bench](https://swe-bench.github.io/) for the evaluation framework and source repositories
290
+ - [Provence](https://arxiv.org/abs/2501.16214) and [SWE-Pruner](https://github.com/ayanami-kitasan/SWE-Pruner) for inspiration on context pruning approaches
chat_template.jinja ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set image_count = namespace(value=0) %}
2
+ {%- set video_count = namespace(value=0) %}
3
+ {%- macro render_content(content, do_vision_count, is_system_content=false) %}
4
+ {%- if content is string %}
5
+ {{- content }}
6
+ {%- elif content is iterable and content is not mapping %}
7
+ {%- for item in content %}
8
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
9
+ {%- if is_system_content %}
10
+ {{- raise_exception('System message cannot contain images.') }}
11
+ {%- endif %}
12
+ {%- if do_vision_count %}
13
+ {%- set image_count.value = image_count.value + 1 %}
14
+ {%- endif %}
15
+ {%- if add_vision_id %}
16
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
17
+ {%- endif %}
18
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
19
+ {%- elif 'video' in item or item.type == 'video' %}
20
+ {%- if is_system_content %}
21
+ {{- raise_exception('System message cannot contain videos.') }}
22
+ {%- endif %}
23
+ {%- if do_vision_count %}
24
+ {%- set video_count.value = video_count.value + 1 %}
25
+ {%- endif %}
26
+ {%- if add_vision_id %}
27
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
28
+ {%- endif %}
29
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
30
+ {%- elif 'text' in item %}
31
+ {{- item.text }}
32
+ {%- else %}
33
+ {{- raise_exception('Unexpected item type in content.') }}
34
+ {%- endif %}
35
+ {%- endfor %}
36
+ {%- elif content is none or content is undefined %}
37
+ {{- '' }}
38
+ {%- else %}
39
+ {{- raise_exception('Unexpected content type.') }}
40
+ {%- endif %}
41
+ {%- endmacro %}
42
+ {%- if not messages %}
43
+ {{- raise_exception('No messages provided.') }}
44
+ {%- endif %}
45
+ {%- if tools and tools is iterable and tools is not mapping %}
46
+ {{- '<|im_start|>system\n' }}
47
+ {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
48
+ {%- for tool in tools %}
49
+ {{- "\n" }}
50
+ {{- tool | tojson }}
51
+ {%- endfor %}
52
+ {{- "\n</tools>" }}
53
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
54
+ {%- if messages[0].role == 'system' %}
55
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
56
+ {%- if content %}
57
+ {{- '\n\n' + content }}
58
+ {%- endif %}
59
+ {%- endif %}
60
+ {{- '<|im_end|>\n' }}
61
+ {%- else %}
62
+ {%- if messages[0].role == 'system' %}
63
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
64
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
65
+ {%- endif %}
66
+ {%- endif %}
67
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
68
+ {%- for message in messages[::-1] %}
69
+ {%- set index = (messages|length - 1) - loop.index0 %}
70
+ {%- if ns.multi_step_tool and message.role == "user" %}
71
+ {%- set content = render_content(message.content, false)|trim %}
72
+ {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
73
+ {%- set ns.multi_step_tool = false %}
74
+ {%- set ns.last_query_index = index %}
75
+ {%- endif %}
76
+ {%- endif %}
77
+ {%- endfor %}
78
+ {%- if ns.multi_step_tool %}
79
+ {{- raise_exception('No user query found in messages.') }}
80
+ {%- endif %}
81
+ {%- for message in messages %}
82
+ {%- set content = render_content(message.content, true)|trim %}
83
+ {%- if message.role == "system" %}
84
+ {%- if not loop.first %}
85
+ {{- raise_exception('System message must be at the beginning.') }}
86
+ {%- endif %}
87
+ {%- elif message.role == "user" %}
88
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
89
+ {%- elif message.role == "assistant" %}
90
+ {%- set reasoning_content = '' %}
91
+ {%- if message.reasoning_content is string %}
92
+ {%- set reasoning_content = message.reasoning_content %}
93
+ {%- else %}
94
+ {%- if '</think>' in content %}
95
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
96
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
97
+ {%- endif %}
98
+ {%- endif %}
99
+ {%- set reasoning_content = reasoning_content|trim %}
100
+ {%- if loop.index0 > ns.last_query_index %}
101
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
102
+ {%- else %}
103
+ {{- '<|im_start|>' + message.role + '\n' + content }}
104
+ {%- endif %}
105
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
106
+ {%- for tool_call in message.tool_calls %}
107
+ {%- if tool_call.function is defined %}
108
+ {%- set tool_call = tool_call.function %}
109
+ {%- endif %}
110
+ {%- if loop.first %}
111
+ {%- if content|trim %}
112
+ {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
113
+ {%- else %}
114
+ {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
115
+ {%- endif %}
116
+ {%- else %}
117
+ {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
118
+ {%- endif %}
119
+ {%- if tool_call.arguments is defined %}
120
+ {%- for args_name, args_value in tool_call.arguments|items %}
121
+ {{- '<parameter=' + args_name + '>\n' }}
122
+ {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
123
+ {{- args_value }}
124
+ {{- '\n</parameter>\n' }}
125
+ {%- endfor %}
126
+ {%- endif %}
127
+ {{- '</function>\n</tool_call>' }}
128
+ {%- endfor %}
129
+ {%- endif %}
130
+ {{- '<|im_end|>\n' }}
131
+ {%- elif message.role == "tool" %}
132
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
133
+ {{- '<|im_start|>user' }}
134
+ {%- endif %}
135
+ {{- '\n<tool_response>\n' }}
136
+ {{- content }}
137
+ {{- '\n</tool_response>' }}
138
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
139
+ {{- '<|im_end|>\n' }}
140
+ {%- elif loop.last %}
141
+ {{- '<|im_end|>\n' }}
142
+ {%- endif %}
143
+ {%- else %}
144
+ {{- raise_exception('Unexpected message role.') }}
145
+ {%- endif %}
146
+ {%- endfor %}
147
+ {%- if add_generation_prompt %}
148
+ {{- '<|im_start|>assistant\n' }}
149
+ {%- if enable_thinking is defined and enable_thinking is true %}
150
+ {{- '<think>\n' }}
151
+ {%- else %}
152
+ {{- '<think>\n\n</think>\n\n' }}
153
+ {%- endif %}
154
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3_5ForConditionalGeneration"
4
+ ],
5
+ "torch_dtype": "bfloat16",
6
+ "image_token_id": 248056,
7
+ "model_name": "Qwen/Qwen3.5-2B",
8
+ "model_type": "qwen3_5",
9
+ "pad_token_id": 248044,
10
+ "text_config": {
11
+ "attention_bias": false,
12
+ "attention_dropout": 0.0,
13
+ "attn_output_gate": true,
14
+ "bos_token_id": null,
15
+ "torch_dtype": "bfloat16",
16
+ "eos_token_id": 248044,
17
+ "full_attention_interval": 4,
18
+ "head_dim": 256,
19
+ "hidden_act": "silu",
20
+ "hidden_size": 2048,
21
+ "initializer_range": 0.02,
22
+ "intermediate_size": 6144,
23
+ "layer_types": [
24
+ "linear_attention",
25
+ "linear_attention",
26
+ "linear_attention",
27
+ "full_attention",
28
+ "linear_attention",
29
+ "linear_attention",
30
+ "linear_attention",
31
+ "full_attention",
32
+ "linear_attention",
33
+ "linear_attention",
34
+ "linear_attention",
35
+ "full_attention",
36
+ "linear_attention",
37
+ "linear_attention",
38
+ "linear_attention",
39
+ "full_attention",
40
+ "linear_attention",
41
+ "linear_attention",
42
+ "linear_attention",
43
+ "full_attention",
44
+ "linear_attention",
45
+ "linear_attention",
46
+ "linear_attention",
47
+ "full_attention"
48
+ ],
49
+ "linear_conv_kernel_dim": 4,
50
+ "linear_key_head_dim": 128,
51
+ "linear_num_key_heads": 16,
52
+ "linear_num_value_heads": 16,
53
+ "linear_value_head_dim": 128,
54
+ "mamba_ssm_dtype": "float32",
55
+ "max_position_embeddings": 262144,
56
+ "mlp_only_layers": [],
57
+ "model_type": "qwen3_5_text",
58
+ "mtp_num_hidden_layers": 1,
59
+ "mtp_use_dedicated_embeddings": false,
60
+ "num_attention_heads": 8,
61
+ "num_hidden_layers": 24,
62
+ "num_key_value_heads": 2,
63
+ "pad_token_id": null,
64
+ "partial_rotary_factor": 0.25,
65
+ "rms_norm_eps": 1e-06,
66
+ "rope_parameters": {
67
+ "mrope_interleaved": true,
68
+ "mrope_section": [
69
+ 11,
70
+ 11,
71
+ 10
72
+ ],
73
+ "partial_rotary_factor": 0.25,
74
+ "rope_theta": 10000000,
75
+ "rope_type": "default"
76
+ },
77
+ "tie_word_embeddings": true,
78
+ "use_cache": true,
79
+ "vocab_size": 248320
80
+ },
81
+ "tie_word_embeddings": true,
82
+ "unsloth_version": "2026.3.4",
83
+ "video_token_id": 248057,
84
+ "vision_config": {
85
+ "deepstack_visual_indexes": [],
86
+ "depth": 24,
87
+ "torch_dtype": "bfloat16",
88
+ "hidden_act": "gelu_pytorch_tanh",
89
+ "hidden_size": 1024,
90
+ "in_channels": 3,
91
+ "initializer_range": 0.02,
92
+ "intermediate_size": 4096,
93
+ "model_type": "qwen3_5",
94
+ "num_heads": 16,
95
+ "num_position_embeddings": 2304,
96
+ "out_hidden_size": 2048,
97
+ "patch_size": 16,
98
+ "spatial_merge_size": 2,
99
+ "temporal_patch_size": 2
100
+ },
101
+ "vision_end_token_id": 248054,
102
+ "vision_start_token_id": 248053
103
+ }
model.safetensors-00001-of-00001.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:390a6b651c6ee0cb7e1d10636056cba492154b52d8204a8fa0f05e904930c0bf
3
+ size 4548221488
model.safetensors.index.json ADDED
@@ -0,0 +1,639 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 4548144832
4
+ },
5
+ "weight_map": {
6
+ "model.language_model.embed_tokens.weight": "model.safetensors-00001-of-00001.safetensors",
7
+ "model.visual.merger.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
8
+ "model.language_model.layers.22.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
9
+ "model.language_model.layers.22.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
10
+ "model.language_model.layers.22.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
11
+ "model.language_model.layers.23.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
12
+ "model.language_model.layers.23.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
13
+ "model.language_model.layers.23.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
14
+ "model.language_model.layers.3.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
15
+ "model.language_model.layers.3.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
16
+ "model.language_model.layers.3.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
17
+ "model.language_model.layers.6.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
18
+ "model.language_model.layers.6.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
19
+ "model.language_model.layers.6.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
20
+ "model.language_model.layers.7.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
21
+ "model.language_model.layers.7.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
22
+ "model.language_model.layers.7.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
23
+ "model.language_model.layers.9.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
24
+ "model.language_model.layers.9.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
25
+ "model.language_model.layers.9.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
26
+ "mtp.layers.0.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
27
+ "mtp.layers.0.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
28
+ "mtp.layers.0.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
29
+ "model.language_model.layers.10.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
30
+ "model.language_model.layers.11.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
31
+ "model.language_model.layers.11.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
32
+ "model.language_model.layers.11.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
33
+ "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
34
+ "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
35
+ "model.language_model.layers.4.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
36
+ "model.language_model.layers.4.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
37
+ "model.language_model.layers.4.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
38
+ "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
39
+ "model.language_model.layers.14.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
40
+ "model.language_model.layers.14.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
41
+ "model.language_model.layers.14.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
42
+ "model.language_model.layers.15.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
43
+ "model.language_model.layers.15.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
44
+ "model.language_model.layers.15.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
45
+ "model.language_model.layers.2.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
46
+ "model.language_model.layers.2.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
47
+ "model.language_model.layers.2.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
48
+ "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
49
+ "model.language_model.layers.20.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
50
+ "model.language_model.layers.20.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
51
+ "model.language_model.layers.20.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
52
+ "model.language_model.layers.17.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
53
+ "model.language_model.layers.17.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
54
+ "model.language_model.layers.17.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
55
+ "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
56
+ "model.language_model.layers.18.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
57
+ "model.language_model.layers.18.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
58
+ "model.language_model.layers.18.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
59
+ "model.language_model.layers.12.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
60
+ "model.language_model.layers.12.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
61
+ "model.language_model.layers.12.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
62
+ "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
63
+ "model.language_model.layers.13.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
64
+ "model.language_model.layers.13.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
65
+ "model.language_model.layers.13.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
66
+ "model.language_model.layers.19.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
67
+ "model.language_model.layers.19.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
68
+ "model.language_model.layers.19.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
69
+ "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
70
+ "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
71
+ "model.language_model.layers.1.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
72
+ "model.language_model.layers.1.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
73
+ "model.language_model.layers.1.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
74
+ "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
75
+ "model.language_model.layers.10.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
76
+ "model.language_model.layers.10.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
77
+ "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
78
+ "model.language_model.layers.16.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
79
+ "model.language_model.layers.16.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
80
+ "model.language_model.layers.16.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
81
+ "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
82
+ "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
83
+ "model.language_model.layers.5.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
84
+ "model.language_model.layers.5.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
85
+ "model.language_model.layers.5.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
86
+ "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
87
+ "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
88
+ "model.language_model.layers.21.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
89
+ "model.language_model.layers.21.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
90
+ "model.language_model.layers.21.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
91
+ "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
92
+ "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
93
+ "model.language_model.layers.8.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
94
+ "model.language_model.layers.8.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
95
+ "model.language_model.layers.8.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
96
+ "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
97
+ "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model.safetensors-00001-of-00001.safetensors",
98
+ "model.language_model.layers.0.mlp.down_proj.weight": "model.safetensors-00001-of-00001.safetensors",
99
+ "model.language_model.layers.0.mlp.gate_proj.weight": "model.safetensors-00001-of-00001.safetensors",
100
+ "model.language_model.layers.0.mlp.up_proj.weight": "model.safetensors-00001-of-00001.safetensors",
101
+ "model.language_model.layers.23.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
102
+ "model.language_model.layers.7.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
103
+ "mtp.fc.weight": "model.safetensors-00001-of-00001.safetensors",
104
+ "mtp.layers.0.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
105
+ "model.language_model.layers.11.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
106
+ "model.language_model.layers.3.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
107
+ "model.language_model.layers.15.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
108
+ "model.language_model.layers.19.self_attn.q_proj.weight": "model.safetensors-00001-of-00001.safetensors",
109
+ "model.visual.merger.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
110
+ "model.language_model.layers.23.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
111
+ "model.language_model.layers.3.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
112
+ "model.language_model.layers.6.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
113
+ "model.language_model.layers.7.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
114
+ "mtp.layers.0.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
115
+ "model.language_model.layers.11.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
116
+ "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
117
+ "model.language_model.layers.12.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
118
+ "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
119
+ "model.language_model.layers.4.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
120
+ "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
121
+ "model.language_model.layers.14.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
122
+ "model.language_model.layers.15.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
123
+ "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
124
+ "model.language_model.layers.20.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
125
+ "model.language_model.layers.17.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
126
+ "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
127
+ "model.language_model.layers.18.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
128
+ "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
129
+ "model.language_model.layers.13.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
130
+ "model.language_model.layers.19.self_attn.o_proj.weight": "model.safetensors-00001-of-00001.safetensors",
131
+ "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
132
+ "model.language_model.layers.2.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
133
+ "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
134
+ "model.language_model.layers.1.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
135
+ "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
136
+ "model.language_model.layers.10.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
137
+ "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
138
+ "model.language_model.layers.16.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
139
+ "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
140
+ "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
141
+ "model.language_model.layers.5.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
142
+ "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
143
+ "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
144
+ "model.language_model.layers.21.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
145
+ "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
146
+ "model.language_model.layers.22.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
147
+ "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
148
+ "model.language_model.layers.8.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
149
+ "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
150
+ "model.language_model.layers.9.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
151
+ "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model.safetensors-00001-of-00001.safetensors",
152
+ "model.language_model.layers.0.linear_attn.out_proj.weight": "model.safetensors-00001-of-00001.safetensors",
153
+ "model.visual.blocks.0.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
154
+ "model.visual.blocks.0.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
155
+ "model.visual.blocks.1.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
156
+ "model.visual.blocks.1.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
157
+ "model.visual.blocks.10.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
158
+ "model.visual.blocks.10.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
159
+ "model.visual.blocks.11.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
160
+ "model.visual.blocks.11.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
161
+ "model.visual.blocks.12.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
162
+ "model.visual.blocks.12.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
163
+ "model.visual.blocks.13.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
164
+ "model.visual.blocks.13.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
165
+ "model.visual.blocks.14.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
166
+ "model.visual.blocks.14.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
167
+ "model.visual.blocks.15.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
168
+ "model.visual.blocks.15.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
169
+ "model.visual.blocks.16.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
170
+ "model.visual.blocks.16.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
171
+ "model.visual.blocks.17.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
172
+ "model.visual.blocks.17.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
173
+ "model.visual.blocks.18.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
174
+ "model.visual.blocks.18.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
175
+ "model.visual.blocks.19.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
176
+ "model.visual.blocks.19.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
177
+ "model.visual.blocks.2.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
178
+ "model.visual.blocks.2.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
179
+ "model.visual.blocks.20.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
180
+ "model.visual.blocks.20.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
181
+ "model.visual.blocks.21.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
182
+ "model.visual.blocks.21.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
183
+ "model.visual.blocks.22.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
184
+ "model.visual.blocks.22.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
185
+ "model.visual.blocks.23.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
186
+ "model.visual.blocks.23.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
187
+ "model.visual.blocks.3.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
188
+ "model.visual.blocks.3.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
189
+ "model.visual.blocks.4.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
190
+ "model.visual.blocks.4.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
191
+ "model.visual.blocks.5.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
192
+ "model.visual.blocks.5.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
193
+ "model.visual.blocks.6.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
194
+ "model.visual.blocks.6.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
195
+ "model.visual.blocks.7.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
196
+ "model.visual.blocks.7.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
197
+ "model.visual.blocks.8.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
198
+ "model.visual.blocks.8.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
199
+ "model.visual.blocks.9.mlp.linear_fc1.weight": "model.safetensors-00001-of-00001.safetensors",
200
+ "model.visual.blocks.9.mlp.linear_fc2.weight": "model.safetensors-00001-of-00001.safetensors",
201
+ "model.visual.blocks.0.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
202
+ "model.visual.blocks.1.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
203
+ "model.visual.blocks.10.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
204
+ "model.visual.blocks.11.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
205
+ "model.visual.blocks.12.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
206
+ "model.visual.blocks.13.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
207
+ "model.visual.blocks.14.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
208
+ "model.visual.blocks.15.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
209
+ "model.visual.blocks.16.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
210
+ "model.visual.blocks.17.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
211
+ "model.visual.blocks.18.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
212
+ "model.visual.blocks.19.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
213
+ "model.visual.blocks.2.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
214
+ "model.visual.blocks.20.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
215
+ "model.visual.blocks.21.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
216
+ "model.visual.blocks.22.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
217
+ "model.visual.blocks.23.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
218
+ "model.visual.blocks.3.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
219
+ "model.visual.blocks.4.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
220
+ "model.visual.blocks.5.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
221
+ "model.visual.blocks.6.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
222
+ "model.visual.blocks.7.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
223
+ "model.visual.blocks.8.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
224
+ "model.visual.blocks.9.attn.qkv.weight": "model.safetensors-00001-of-00001.safetensors",
225
+ "model.visual.pos_embed.weight": "model.safetensors-00001-of-00001.safetensors",
226
+ "model.visual.patch_embed.proj.weight": "model.safetensors-00001-of-00001.safetensors",
227
+ "model.language_model.layers.23.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
228
+ "model.language_model.layers.23.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
229
+ "model.language_model.layers.7.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
230
+ "model.language_model.layers.7.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
231
+ "mtp.layers.0.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
232
+ "mtp.layers.0.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
233
+ "model.language_model.layers.11.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
234
+ "model.language_model.layers.11.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
235
+ "model.language_model.layers.3.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
236
+ "model.language_model.layers.3.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
237
+ "model.language_model.layers.15.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
238
+ "model.language_model.layers.15.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
239
+ "model.language_model.layers.19.self_attn.k_proj.weight": "model.safetensors-00001-of-00001.safetensors",
240
+ "model.language_model.layers.19.self_attn.v_proj.weight": "model.safetensors-00001-of-00001.safetensors",
241
+ "model.visual.blocks.0.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
242
+ "model.visual.blocks.1.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
243
+ "model.visual.blocks.10.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
244
+ "model.visual.blocks.11.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
245
+ "model.visual.blocks.12.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
246
+ "model.visual.blocks.13.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
247
+ "model.visual.blocks.14.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
248
+ "model.visual.blocks.15.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
249
+ "model.visual.blocks.16.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
250
+ "model.visual.blocks.17.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
251
+ "model.visual.blocks.18.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
252
+ "model.visual.blocks.19.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
253
+ "model.visual.blocks.2.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
254
+ "model.visual.blocks.20.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
255
+ "model.visual.blocks.21.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
256
+ "model.visual.blocks.22.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
257
+ "model.visual.blocks.23.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
258
+ "model.visual.blocks.3.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
259
+ "model.visual.blocks.4.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
260
+ "model.visual.blocks.5.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
261
+ "model.visual.blocks.6.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
262
+ "model.visual.blocks.7.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
263
+ "model.visual.blocks.8.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
264
+ "model.visual.blocks.9.attn.proj.weight": "model.safetensors-00001-of-00001.safetensors",
265
+ "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
266
+ "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
267
+ "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
268
+ "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
269
+ "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
270
+ "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
271
+ "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
272
+ "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
273
+ "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
274
+ "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
275
+ "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
276
+ "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
277
+ "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
278
+ "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
279
+ "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
280
+ "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
281
+ "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
282
+ "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
283
+ "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
284
+ "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
285
+ "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
286
+ "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
287
+ "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
288
+ "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
289
+ "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
290
+ "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
291
+ "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
292
+ "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
293
+ "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
294
+ "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
295
+ "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
296
+ "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
297
+ "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
298
+ "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
299
+ "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model.safetensors-00001-of-00001.safetensors",
300
+ "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model.safetensors-00001-of-00001.safetensors",
301
+ "model.language_model.layers.8.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
302
+ "model.language_model.layers.12.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
303
+ "model.language_model.layers.4.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
304
+ "model.language_model.layers.5.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
305
+ "model.language_model.layers.20.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
306
+ "model.language_model.layers.21.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
307
+ "model.language_model.layers.18.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
308
+ "model.language_model.layers.13.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
309
+ "model.language_model.layers.14.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
310
+ "model.language_model.layers.2.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
311
+ "model.language_model.layers.10.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
312
+ "model.language_model.layers.16.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
313
+ "model.language_model.layers.17.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
314
+ "model.language_model.layers.6.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
315
+ "model.language_model.layers.22.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
316
+ "model.language_model.layers.9.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
317
+ "model.language_model.layers.0.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
318
+ "model.language_model.layers.1.linear_attn.conv1d.weight": "model.safetensors-00001-of-00001.safetensors",
319
+ "model.visual.blocks.0.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
320
+ "model.visual.blocks.1.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
321
+ "model.visual.blocks.10.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
322
+ "model.visual.blocks.11.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
323
+ "model.visual.blocks.12.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
324
+ "model.visual.blocks.13.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
325
+ "model.visual.blocks.14.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
326
+ "model.visual.blocks.15.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
327
+ "model.visual.blocks.16.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
328
+ "model.visual.blocks.17.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
329
+ "model.visual.blocks.18.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
330
+ "model.visual.blocks.19.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
331
+ "model.visual.blocks.2.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
332
+ "model.visual.blocks.20.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
333
+ "model.visual.blocks.21.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
334
+ "model.visual.blocks.22.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
335
+ "model.visual.blocks.23.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
336
+ "model.visual.blocks.3.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
337
+ "model.visual.blocks.4.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
338
+ "model.visual.blocks.5.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
339
+ "model.visual.blocks.6.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
340
+ "model.visual.blocks.7.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
341
+ "model.visual.blocks.8.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
342
+ "model.visual.blocks.9.mlp.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
343
+ "model.visual.merger.linear_fc1.bias": "model.safetensors-00001-of-00001.safetensors",
344
+ "model.visual.blocks.0.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
345
+ "model.visual.blocks.1.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
346
+ "model.visual.blocks.10.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
347
+ "model.visual.blocks.11.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
348
+ "model.visual.blocks.12.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
349
+ "model.visual.blocks.13.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
350
+ "model.visual.blocks.14.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
351
+ "model.visual.blocks.15.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
352
+ "model.visual.blocks.16.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
353
+ "model.visual.blocks.17.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
354
+ "model.visual.blocks.18.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
355
+ "model.visual.blocks.19.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
356
+ "model.visual.blocks.2.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
357
+ "model.visual.blocks.20.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
358
+ "model.visual.blocks.21.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
359
+ "model.visual.blocks.22.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
360
+ "model.visual.blocks.23.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
361
+ "model.visual.blocks.3.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
362
+ "model.visual.blocks.4.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
363
+ "model.visual.blocks.5.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
364
+ "model.visual.blocks.6.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
365
+ "model.visual.blocks.7.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
366
+ "model.visual.blocks.8.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
367
+ "model.visual.blocks.9.attn.qkv.bias": "model.safetensors-00001-of-00001.safetensors",
368
+ "model.language_model.layers.22.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
369
+ "model.language_model.layers.23.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
370
+ "model.language_model.layers.23.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
371
+ "model.language_model.layers.3.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
372
+ "model.language_model.layers.6.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
373
+ "model.language_model.layers.7.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
374
+ "model.language_model.layers.7.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
375
+ "model.language_model.layers.8.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
376
+ "mtp.layers.0.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
377
+ "mtp.layers.0.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
378
+ "mtp.norm.weight": "model.safetensors-00001-of-00001.safetensors",
379
+ "mtp.pre_fc_norm_embedding.weight": "model.safetensors-00001-of-00001.safetensors",
380
+ "mtp.pre_fc_norm_hidden.weight": "model.safetensors-00001-of-00001.safetensors",
381
+ "model.language_model.layers.11.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
382
+ "model.language_model.layers.11.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
383
+ "model.language_model.layers.12.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
384
+ "model.language_model.layers.3.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
385
+ "model.language_model.layers.4.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
386
+ "model.language_model.layers.4.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
387
+ "model.language_model.layers.14.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
388
+ "model.language_model.layers.14.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
389
+ "model.language_model.layers.15.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
390
+ "model.language_model.layers.15.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
391
+ "model.language_model.layers.20.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
392
+ "model.language_model.layers.20.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
393
+ "model.language_model.layers.17.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
394
+ "model.language_model.layers.18.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
395
+ "model.language_model.layers.18.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
396
+ "model.language_model.layers.19.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
397
+ "model.language_model.layers.12.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
398
+ "model.language_model.layers.13.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
399
+ "model.language_model.layers.13.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
400
+ "model.language_model.layers.19.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
401
+ "model.language_model.layers.2.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
402
+ "model.language_model.layers.2.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
403
+ "model.language_model.layers.1.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
404
+ "model.language_model.layers.10.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
405
+ "model.language_model.layers.10.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
406
+ "model.language_model.layers.16.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
407
+ "model.language_model.layers.16.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
408
+ "model.language_model.layers.17.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
409
+ "model.language_model.layers.5.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
410
+ "model.language_model.layers.5.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
411
+ "model.language_model.layers.6.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
412
+ "model.language_model.layers.21.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
413
+ "model.language_model.layers.21.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
414
+ "model.language_model.layers.22.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
415
+ "model.language_model.layers.8.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
416
+ "model.language_model.layers.9.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
417
+ "model.language_model.layers.9.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
418
+ "model.language_model.layers.0.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
419
+ "model.language_model.layers.0.post_attention_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
420
+ "model.language_model.layers.1.input_layernorm.weight": "model.safetensors-00001-of-00001.safetensors",
421
+ "model.language_model.norm.weight": "model.safetensors-00001-of-00001.safetensors",
422
+ "model.visual.merger.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
423
+ "model.visual.blocks.0.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
424
+ "model.visual.blocks.0.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
425
+ "model.visual.blocks.0.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
426
+ "model.visual.blocks.0.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
427
+ "model.visual.blocks.0.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
428
+ "model.visual.blocks.0.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
429
+ "model.visual.blocks.1.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
430
+ "model.visual.blocks.1.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
431
+ "model.visual.blocks.1.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
432
+ "model.visual.blocks.1.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
433
+ "model.visual.blocks.1.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
434
+ "model.visual.blocks.1.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
435
+ "model.visual.blocks.10.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
436
+ "model.visual.blocks.10.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
437
+ "model.visual.blocks.10.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
438
+ "model.visual.blocks.10.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
439
+ "model.visual.blocks.10.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
440
+ "model.visual.blocks.10.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
441
+ "model.visual.blocks.11.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
442
+ "model.visual.blocks.11.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
443
+ "model.visual.blocks.11.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
444
+ "model.visual.blocks.11.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
445
+ "model.visual.blocks.11.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
446
+ "model.visual.blocks.11.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
447
+ "model.visual.blocks.12.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
448
+ "model.visual.blocks.12.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
449
+ "model.visual.blocks.12.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
450
+ "model.visual.blocks.12.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
451
+ "model.visual.blocks.12.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
452
+ "model.visual.blocks.12.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
453
+ "model.visual.blocks.13.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
454
+ "model.visual.blocks.13.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
455
+ "model.visual.blocks.13.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
456
+ "model.visual.blocks.13.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
457
+ "model.visual.blocks.13.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
458
+ "model.visual.blocks.13.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
459
+ "model.visual.blocks.14.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
460
+ "model.visual.blocks.14.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
461
+ "model.visual.blocks.14.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
462
+ "model.visual.blocks.14.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
463
+ "model.visual.blocks.14.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
464
+ "model.visual.blocks.14.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
465
+ "model.visual.blocks.15.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
466
+ "model.visual.blocks.15.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
467
+ "model.visual.blocks.15.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
468
+ "model.visual.blocks.15.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
469
+ "model.visual.blocks.15.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
470
+ "model.visual.blocks.15.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
471
+ "model.visual.blocks.16.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
472
+ "model.visual.blocks.16.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
473
+ "model.visual.blocks.16.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
474
+ "model.visual.blocks.16.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
475
+ "model.visual.blocks.16.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
476
+ "model.visual.blocks.16.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
477
+ "model.visual.blocks.17.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
478
+ "model.visual.blocks.17.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
479
+ "model.visual.blocks.17.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
480
+ "model.visual.blocks.17.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
481
+ "model.visual.blocks.17.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
482
+ "model.visual.blocks.17.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
483
+ "model.visual.blocks.18.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
484
+ "model.visual.blocks.18.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
485
+ "model.visual.blocks.18.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
486
+ "model.visual.blocks.18.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
487
+ "model.visual.blocks.18.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
488
+ "model.visual.blocks.18.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
489
+ "model.visual.blocks.19.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
490
+ "model.visual.blocks.19.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
491
+ "model.visual.blocks.19.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
492
+ "model.visual.blocks.19.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
493
+ "model.visual.blocks.19.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
494
+ "model.visual.blocks.19.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
495
+ "model.visual.blocks.2.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
496
+ "model.visual.blocks.2.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
497
+ "model.visual.blocks.2.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
498
+ "model.visual.blocks.2.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
499
+ "model.visual.blocks.2.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
500
+ "model.visual.blocks.2.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
501
+ "model.visual.blocks.20.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
502
+ "model.visual.blocks.20.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
503
+ "model.visual.blocks.20.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
504
+ "model.visual.blocks.20.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
505
+ "model.visual.blocks.20.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
506
+ "model.visual.blocks.20.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
507
+ "model.visual.blocks.21.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
508
+ "model.visual.blocks.21.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
509
+ "model.visual.blocks.21.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
510
+ "model.visual.blocks.21.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
511
+ "model.visual.blocks.21.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
512
+ "model.visual.blocks.21.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
513
+ "model.visual.blocks.22.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
514
+ "model.visual.blocks.22.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
515
+ "model.visual.blocks.22.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
516
+ "model.visual.blocks.22.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
517
+ "model.visual.blocks.22.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
518
+ "model.visual.blocks.22.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
519
+ "model.visual.blocks.23.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
520
+ "model.visual.blocks.23.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
521
+ "model.visual.blocks.23.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
522
+ "model.visual.blocks.23.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
523
+ "model.visual.blocks.23.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
524
+ "model.visual.blocks.23.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
525
+ "model.visual.blocks.3.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
526
+ "model.visual.blocks.3.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
527
+ "model.visual.blocks.3.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
528
+ "model.visual.blocks.3.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
529
+ "model.visual.blocks.3.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
530
+ "model.visual.blocks.3.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
531
+ "model.visual.blocks.4.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
532
+ "model.visual.blocks.4.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
533
+ "model.visual.blocks.4.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
534
+ "model.visual.blocks.4.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
535
+ "model.visual.blocks.4.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
536
+ "model.visual.blocks.4.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
537
+ "model.visual.blocks.5.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
538
+ "model.visual.blocks.5.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
539
+ "model.visual.blocks.5.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
540
+ "model.visual.blocks.5.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
541
+ "model.visual.blocks.5.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
542
+ "model.visual.blocks.5.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
543
+ "model.visual.blocks.6.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
544
+ "model.visual.blocks.6.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
545
+ "model.visual.blocks.6.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
546
+ "model.visual.blocks.6.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
547
+ "model.visual.blocks.6.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
548
+ "model.visual.blocks.6.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
549
+ "model.visual.blocks.7.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
550
+ "model.visual.blocks.7.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
551
+ "model.visual.blocks.7.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
552
+ "model.visual.blocks.7.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
553
+ "model.visual.blocks.7.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
554
+ "model.visual.blocks.7.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
555
+ "model.visual.blocks.8.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
556
+ "model.visual.blocks.8.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
557
+ "model.visual.blocks.8.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
558
+ "model.visual.blocks.8.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
559
+ "model.visual.blocks.8.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
560
+ "model.visual.blocks.8.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
561
+ "model.visual.blocks.9.attn.proj.bias": "model.safetensors-00001-of-00001.safetensors",
562
+ "model.visual.blocks.9.mlp.linear_fc2.bias": "model.safetensors-00001-of-00001.safetensors",
563
+ "model.visual.blocks.9.norm1.bias": "model.safetensors-00001-of-00001.safetensors",
564
+ "model.visual.blocks.9.norm1.weight": "model.safetensors-00001-of-00001.safetensors",
565
+ "model.visual.blocks.9.norm2.bias": "model.safetensors-00001-of-00001.safetensors",
566
+ "model.visual.blocks.9.norm2.weight": "model.safetensors-00001-of-00001.safetensors",
567
+ "model.visual.merger.norm.bias": "model.safetensors-00001-of-00001.safetensors",
568
+ "model.visual.merger.norm.weight": "model.safetensors-00001-of-00001.safetensors",
569
+ "model.visual.patch_embed.proj.bias": "model.safetensors-00001-of-00001.safetensors",
570
+ "model.language_model.layers.23.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
571
+ "model.language_model.layers.23.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
572
+ "model.language_model.layers.3.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
573
+ "model.language_model.layers.7.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
574
+ "model.language_model.layers.7.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
575
+ "mtp.layers.0.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
576
+ "mtp.layers.0.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
577
+ "model.language_model.layers.11.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
578
+ "model.language_model.layers.11.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
579
+ "model.language_model.layers.12.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
580
+ "model.language_model.layers.3.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
581
+ "model.language_model.layers.4.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
582
+ "model.language_model.layers.14.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
583
+ "model.language_model.layers.15.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
584
+ "model.language_model.layers.15.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
585
+ "model.language_model.layers.20.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
586
+ "model.language_model.layers.18.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
587
+ "model.language_model.layers.13.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
588
+ "model.language_model.layers.19.self_attn.k_norm.weight": "model.safetensors-00001-of-00001.safetensors",
589
+ "model.language_model.layers.19.self_attn.q_norm.weight": "model.safetensors-00001-of-00001.safetensors",
590
+ "model.language_model.layers.2.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
591
+ "model.language_model.layers.1.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
592
+ "model.language_model.layers.10.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
593
+ "model.language_model.layers.16.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
594
+ "model.language_model.layers.17.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
595
+ "model.language_model.layers.5.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
596
+ "model.language_model.layers.6.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
597
+ "model.language_model.layers.21.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
598
+ "model.language_model.layers.22.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
599
+ "model.language_model.layers.8.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
600
+ "model.language_model.layers.9.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
601
+ "model.language_model.layers.0.linear_attn.norm.weight": "model.safetensors-00001-of-00001.safetensors",
602
+ "model.language_model.layers.8.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
603
+ "model.language_model.layers.12.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
604
+ "model.language_model.layers.4.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
605
+ "model.language_model.layers.5.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
606
+ "model.language_model.layers.16.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
607
+ "model.language_model.layers.20.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
608
+ "model.language_model.layers.21.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
609
+ "model.language_model.layers.18.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
610
+ "model.language_model.layers.13.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
611
+ "model.language_model.layers.14.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
612
+ "model.language_model.layers.2.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
613
+ "model.language_model.layers.10.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
614
+ "model.language_model.layers.17.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
615
+ "model.language_model.layers.6.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
616
+ "model.language_model.layers.22.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
617
+ "model.language_model.layers.9.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
618
+ "model.language_model.layers.0.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
619
+ "model.language_model.layers.1.linear_attn.A_log": "model.safetensors-00001-of-00001.safetensors",
620
+ "model.language_model.layers.8.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
621
+ "model.language_model.layers.12.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
622
+ "model.language_model.layers.4.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
623
+ "model.language_model.layers.20.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
624
+ "model.language_model.layers.21.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
625
+ "model.language_model.layers.18.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
626
+ "model.language_model.layers.13.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
627
+ "model.language_model.layers.14.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
628
+ "model.language_model.layers.2.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
629
+ "model.language_model.layers.10.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
630
+ "model.language_model.layers.16.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
631
+ "model.language_model.layers.17.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
632
+ "model.language_model.layers.5.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
633
+ "model.language_model.layers.6.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
634
+ "model.language_model.layers.22.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
635
+ "model.language_model.layers.9.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
636
+ "model.language_model.layers.0.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors",
637
+ "model.language_model.layers.1.linear_attn.dt_bias": "model.safetensors-00001-of-00001.safetensors"
638
+ }
639
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {
3
+ "longest_edge": 16777216,
4
+ "shortest_edge": 65536
5
+ },
6
+ "patch_size": 16,
7
+ "temporal_patch_size": 2,
8
+ "merge_size": 2,
9
+ "image_mean": [
10
+ 0.5,
11
+ 0.5,
12
+ 0.5
13
+ ],
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "processor_class": "Qwen3VLProcessor",
20
+ "image_processor_type": "Qwen2VLImageProcessorFast"
21
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
3
+ size 19989343
tokenizer_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "audio_bos_token": "<|audio_start|>",
4
+ "audio_eos_token": "<|audio_end|>",
5
+ "audio_token": "<|audio_pad|>",
6
+ "backend": "tokenizers",
7
+ "bos_token": null,
8
+ "clean_up_tokenization_spaces": false,
9
+ "eos_token": "<|im_end|>",
10
+ "errors": "replace",
11
+ "image_token": "<|image_pad|>",
12
+ "is_local": false,
13
+ "model_max_length": 262144,
14
+ "model_specific_special_tokens": {
15
+ "audio_bos_token": "<|audio_start|>",
16
+ "audio_eos_token": "<|audio_end|>",
17
+ "audio_token": "<|audio_pad|>",
18
+ "image_token": "<|image_pad|>",
19
+ "video_token": "<|video_pad|>",
20
+ "vision_bos_token": "<|vision_start|>",
21
+ "vision_eos_token": "<|vision_end|>"
22
+ },
23
+ "pad_token": "<|endoftext|>",
24
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
25
+ "split_special_tokens": false,
26
+ "tokenizer_class": "TokenizersBackend",
27
+ "unk_token": null,
28
+ "video_token": "<|video_pad|>",
29
+ "vision_bos_token": "<|vision_start|>",
30
+ "vision_eos_token": "<|vision_end|>",
31
+ "chat_template": "{%- set image_count = namespace(value=0) %}\n{%- set video_count = namespace(value=0) %}\n{%- macro render_content(content, do_vision_count, is_system_content=false) %}\n {%- if content is string %}\n {{- content }}\n {%- elif content is iterable and content is not mapping %}\n {%- for item in content %}\n {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}\n {%- if is_system_content %}\n {{- raise_exception('System message cannot contain images.') }}\n {%- endif %}\n {%- if do_vision_count %}\n {%- set image_count.value = image_count.value + 1 %}\n {%- endif %}\n {%- if add_vision_id %}\n {{- 'Picture ' ~ image_count.value ~ ': ' }}\n {%- endif %}\n {{- '<|vision_start|><|image_pad|><|vision_end|>' }}\n {%- elif 'video' in item or item.type == 'video' %}\n {%- if is_system_content %}\n {{- raise_exception('System message cannot contain videos.') }}\n {%- endif %}\n {%- if do_vision_count %}\n {%- set video_count.value = video_count.value + 1 %}\n {%- endif %}\n {%- if add_vision_id %}\n {{- 'Video ' ~ video_count.value ~ ': ' }}\n {%- endif %}\n {{- '<|vision_start|><|video_pad|><|vision_end|>' }}\n {%- elif 'text' in item %}\n {{- item.text }}\n {%- else %}\n {{- raise_exception('Unexpected item type in content.') }}\n {%- endif %}\n {%- endfor %}\n {%- elif content is none or content is undefined %}\n {{- '' }}\n {%- else %}\n {{- raise_exception('Unexpected content type.') }}\n {%- endif %}\n{%- endmacro %}\n{%- if not messages %}\n {{- raise_exception('No messages provided.') }}\n{%- endif %}\n{%- if tools and tools is iterable and tools is not mapping %}\n {{- '<|im_start|>system\\n' }}\n {{- \"# Tools\\n\\nYou have access to the following functions:\\n\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\" }}\n {{- '\\n\\nIf you choose to call a function ONLY reply in the following format with NO suffix:\\n\\n<tool_call>\\n<function=example_function_name>\\n<parameter=example_parameter_1>\\nvalue_1\\n</parameter>\\n<parameter=example_parameter_2>\\nThis is the value for the second parameter\\nthat can span\\nmultiple lines\\n</parameter>\\n</function>\\n</tool_call>\\n\\n<IMPORTANT>\\nReminder:\\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\\n- Required parameters MUST be specified\\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\\n</IMPORTANT>' }}\n {%- if messages[0].role == 'system' %}\n {%- set content = render_content(messages[0].content, false, true)|trim %}\n {%- if content %}\n {{- '\\n\\n' + content }}\n {%- endif %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {%- set content = render_content(messages[0].content, false, true)|trim %}\n {{- '<|im_start|>system\\n' + content + '<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n {%- set index = (messages|length - 1) - loop.index0 %}\n {%- if ns.multi_step_tool and message.role == \"user\" %}\n {%- set content = render_content(message.content, false)|trim %}\n {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if ns.multi_step_tool %}\n {{- raise_exception('No user query found in messages.') }}\n{%- endif %}\n{%- for message in messages %}\n {%- set content = render_content(message.content, true)|trim %}\n {%- if message.role == \"system\" %}\n {%- if not loop.first %}\n {{- raise_exception('System message must be at the beginning.') }}\n {%- endif %}\n {%- elif message.role == \"user\" %}\n {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {%- set reasoning_content = '' %}\n {%- if message.reasoning_content is string %}\n {%- set reasoning_content = message.reasoning_content %}\n {%- else %}\n {%- if '</think>' in content %}\n {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n {%- endif %}\n {%- endif %}\n {%- set reasoning_content = reasoning_content|trim %}\n {%- if loop.index0 > ns.last_query_index %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content + '\\n</think>\\n\\n' + content }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {%- if loop.first %}\n {%- if content|trim %}\n {{- '\\n\\n<tool_call>\\n<function=' + tool_call.name + '>\\n' }}\n {%- else %}\n {{- '<tool_call>\\n<function=' + tool_call.name + '>\\n' }}\n {%- endif %}\n {%- else %}\n {{- '\\n<tool_call>\\n<function=' + tool_call.name + '>\\n' }}\n {%- endif %}\n {%- if tool_call.arguments is defined %}\n {%- for args_name, args_value in tool_call.arguments|items %}\n {{- '<parameter=' + args_name + '>\\n' }}\n {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}\n {{- args_value }}\n {{- '\\n</parameter>\\n' }}\n {%- endfor %}\n {%- endif %}\n {{- '</function>\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.previtem and loop.previtem.role != \"tool\" %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- content }}\n {{- '\\n</tool_response>' }}\n {%- if not loop.last and loop.nextitem.role != \"tool\" %}\n {{- '<|im_end|>\\n' }}\n {%- elif loop.last %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- else %}\n {{- raise_exception('Unexpected message role.') }}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is true %}\n {{- '<think>\\n' }}\n {%- else %}\n {{- '<think>\\n\\n</think>\\n\\n' }}\n {%- endif %}\n{%- endif %}"
32
+ }