kepeng commited on
Commit
02de2c8
·
verified ·
1 Parent(s): 230737c

Upload folder using huggingface_hub

Browse files
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen3-VL-2B-Instruct
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:Qwen/Qwen3-VL-2B-Instruct
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.17.1
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/adapter_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "Qwen/Qwen3-VL-2B-Instruct",
5
+ "bias": "none",
6
+ "corda_config": null,
7
+ "eva_config": null,
8
+ "exclude_modules": null,
9
+ "fan_in_fan_out": false,
10
+ "inference_mode": true,
11
+ "init_lora_weights": true,
12
+ "layer_replication": null,
13
+ "layers_pattern": null,
14
+ "layers_to_transform": null,
15
+ "loftq_config": {},
16
+ "lora_alpha": 32,
17
+ "lora_bias": false,
18
+ "lora_dropout": 0.05,
19
+ "megatron_config": null,
20
+ "megatron_core": "megatron.core",
21
+ "modules_to_save": [],
22
+ "peft_type": "LORA",
23
+ "qalora_group_size": 16,
24
+ "r": 8,
25
+ "rank_pattern": {},
26
+ "revision": null,
27
+ "target_modules": "^(model.language_model.*\\.(down_proj|up_proj|v_proj|gate_proj|o_proj|q_proj|k_proj)|(?!(model.visual.merger|model.visual.deepstack_merger_list))model.visual.*\\.(camera_metadata_proj.0|linear_fc1|qkv|linear_fc2|attn.proj|camera_metadata_proj.2))$",
28
+ "target_parameters": null,
29
+ "task_type": "CAUSAL_LM",
30
+ "trainable_token_indices": null,
31
+ "use_dora": false,
32
+ "use_qalora": false,
33
+ "use_rslora": false
34
+ }
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c29fc75f5474c24276c2dbc1c4fc2c1f8bc90648451982d86c0a3c41801955d
3
+ size 47630536
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/additional_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/args.json ADDED
@@ -0,0 +1,366 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "<release-local-path-redacted>",
3
+ "overwrite_output_dir": false,
4
+ "do_train": false,
5
+ "do_eval": false,
6
+ "do_predict": false,
7
+ "eval_strategy": "steps",
8
+ "prediction_loss_only": false,
9
+ "per_device_train_batch_size": 1,
10
+ "per_device_eval_batch_size": 2,
11
+ "per_gpu_train_batch_size": null,
12
+ "per_gpu_eval_batch_size": null,
13
+ "gradient_accumulation_steps": 4,
14
+ "eval_accumulation_steps": null,
15
+ "eval_delay": 0,
16
+ "torch_empty_cache_steps": null,
17
+ "learning_rate": 0.0001,
18
+ "weight_decay": 0.1,
19
+ "adam_beta1": 0.9,
20
+ "adam_beta2": 0.95,
21
+ "adam_epsilon": 1e-08,
22
+ "max_grad_norm": 1.0,
23
+ "num_train_epochs": 3.0,
24
+ "max_steps": 100000,
25
+ "lr_scheduler_type": "cosine",
26
+ "lr_scheduler_kwargs": null,
27
+ "warmup_ratio": 0.05,
28
+ "warmup_steps": 0,
29
+ "log_level": "passive",
30
+ "log_level_replica": "warning",
31
+ "log_on_each_node": true,
32
+ "logging_dir": "<release-local-path-redacted>",
33
+ "logging_strategy": "steps",
34
+ "logging_first_step": true,
35
+ "logging_steps": 5,
36
+ "logging_nan_inf_filter": true,
37
+ "save_strategy": "steps",
38
+ "save_steps": 5000.0,
39
+ "save_total_limit": 20,
40
+ "save_safetensors": true,
41
+ "save_on_each_node": false,
42
+ "save_only_model": false,
43
+ "restore_callback_states_from_checkpoint": false,
44
+ "no_cuda": false,
45
+ "use_cpu": false,
46
+ "use_mps_device": false,
47
+ "seed": 42,
48
+ "data_seed": 42,
49
+ "jit_mode_eval": false,
50
+ "bf16": true,
51
+ "fp16": false,
52
+ "fp16_opt_level": "O1",
53
+ "half_precision_backend": "auto",
54
+ "bf16_full_eval": false,
55
+ "fp16_full_eval": false,
56
+ "tf32": null,
57
+ "local_rank": 0,
58
+ "ddp_backend": null,
59
+ "tpu_num_cores": null,
60
+ "tpu_metrics_debug": false,
61
+ "debug": null,
62
+ "dataloader_drop_last": false,
63
+ "eval_steps": 5000.0,
64
+ "dataloader_num_workers": 8,
65
+ "dataloader_prefetch_factor": null,
66
+ "past_index": -1,
67
+ "run_name": "<release-local-path-redacted>",
68
+ "disable_tqdm": null,
69
+ "remove_unused_columns": false,
70
+ "label_names": null,
71
+ "load_best_model_at_end": false,
72
+ "metric_for_best_model": "loss",
73
+ "greater_is_better": false,
74
+ "ignore_data_skip": false,
75
+ "fsdp": [],
76
+ "fsdp_min_num_params": 0,
77
+ "fsdp_config": null,
78
+ "fsdp_transformer_layer_cls_to_wrap": null,
79
+ "accelerator_config": {
80
+ "dispatch_batches": false
81
+ },
82
+ "parallelism_config": null,
83
+ "deepspeed": null,
84
+ "label_smoothing_factor": 0.0,
85
+ "optim": "adamw_torch_fused",
86
+ "optim_args": null,
87
+ "adafactor": false,
88
+ "group_by_length": false,
89
+ "group_by_size": true,
90
+ "length_column_name": "length",
91
+ "report_to": [
92
+ "tensorboard"
93
+ ],
94
+ "project": "huggingface",
95
+ "trackio_space_id": "trackio",
96
+ "ddp_find_unused_parameters": null,
97
+ "ddp_bucket_cap_mb": null,
98
+ "ddp_broadcast_buffers": null,
99
+ "dataloader_pin_memory": true,
100
+ "dataloader_persistent_workers": false,
101
+ "skip_memory_metrics": true,
102
+ "use_legacy_prediction_loop": false,
103
+ "push_to_hub": false,
104
+ "resume_from_checkpoint": "<release-local-path-redacted>",
105
+ "hub_model_id": null,
106
+ "hub_strategy": "every_save",
107
+ "hub_token": null,
108
+ "hub_private_repo": null,
109
+ "hub_always_push": false,
110
+ "hub_revision": null,
111
+ "gradient_checkpointing": true,
112
+ "gradient_checkpointing_kwargs": null,
113
+ "include_inputs_for_metrics": false,
114
+ "include_for_metrics": [],
115
+ "eval_do_concat_batches": true,
116
+ "fp16_backend": "auto",
117
+ "push_to_hub_model_id": null,
118
+ "push_to_hub_organization": null,
119
+ "push_to_hub_token": null,
120
+ "mp_parameters": "",
121
+ "auto_find_batch_size": false,
122
+ "full_determinism": false,
123
+ "torchdynamo": null,
124
+ "ray_scope": "last",
125
+ "ddp_timeout": 18000000,
126
+ "torch_compile": false,
127
+ "torch_compile_backend": null,
128
+ "torch_compile_mode": null,
129
+ "include_tokens_per_second": false,
130
+ "include_num_input_tokens_seen": false,
131
+ "neftune_noise_alpha": null,
132
+ "optim_target_modules": null,
133
+ "batch_eval_metrics": false,
134
+ "eval_on_start": false,
135
+ "use_liger_kernel": false,
136
+ "liger_kernel_config": null,
137
+ "eval_use_gather_object": false,
138
+ "average_tokens_across_devices": true,
139
+ "sortish_sampler": false,
140
+ "predict_with_generate": false,
141
+ "generation_max_length": null,
142
+ "generation_num_beams": null,
143
+ "generation_config": null,
144
+ "tuner_backend": "peft",
145
+ "vit_gradient_checkpointing": false,
146
+ "router_aux_loss_coef": 0.0,
147
+ "enable_dft_loss": false,
148
+ "enable_channel_loss": false,
149
+ "check_model": true,
150
+ "acc_strategy": "token",
151
+ "train_dataloader_shuffle": true,
152
+ "max_epochs": null,
153
+ "aligner_lr": null,
154
+ "vit_lr": null,
155
+ "use_logits_to_keep": null,
156
+ "ds3_gather_for_generation": true,
157
+ "resume_only_model": false,
158
+ "optimizer": null,
159
+ "loss_type": null,
160
+ "metric": null,
161
+ "eval_use_evalscope": false,
162
+ "eval_dataset": [],
163
+ "eval_dataset_args": null,
164
+ "eval_limit": null,
165
+ "eval_generation_config": null,
166
+ "extra_eval_args": null,
167
+ "use_flash_ckpt": false,
168
+ "use_ray": false,
169
+ "ray_exp_name": null,
170
+ "device_groups": null,
171
+ "model": "Qwen/Qwen3-VL-2B-Instruct",
172
+ "model_type": "qwen3_vl",
173
+ "model_revision": null,
174
+ "task_type": "causal_lm",
175
+ "torch_dtype": "bfloat16",
176
+ "attn_impl": "flash_attn",
177
+ "new_special_tokens": [],
178
+ "num_labels": null,
179
+ "problem_type": null,
180
+ "rope_scaling": null,
181
+ "device_map": null,
182
+ "max_memory": {},
183
+ "max_model_len": null,
184
+ "local_repo_path": null,
185
+ "init_strategy": null,
186
+ "camera_metadata_residual_enabled": true,
187
+ "camera_metadata_residual_num_layers": 2,
188
+ "camera_metadata_require_all": false,
189
+ "camera_metadata_norm_scheme": "photographic_log",
190
+ "template": "qwen3_vl",
191
+ "system": null,
192
+ "max_length": 5120,
193
+ "truncation_strategy": "delete",
194
+ "max_pixels": null,
195
+ "agent_template": null,
196
+ "norm_bbox": null,
197
+ "use_chat_template": true,
198
+ "padding_side": "right",
199
+ "padding_free": false,
200
+ "loss_scale": "default",
201
+ "sequence_parallel_size": 1,
202
+ "template_backend": "swift",
203
+ "response_prefix": null,
204
+ "enable_thinking": null,
205
+ "add_non_thinking_prefix": true,
206
+ "dataset": [
207
+ "KEPENGXU/TRAIN_BALANCE_150K_META",
208
+ "KEPENGXU/TRAIN_BALANCE_150K_META_EXTRA"
209
+ ],
210
+ "val_dataset": [
211
+ "KEPENGXU/TEST_RAW_FULL"
212
+ ],
213
+ "cached_dataset": [],
214
+ "cached_val_dataset": [],
215
+ "split_dataset_ratio": 0.0,
216
+ "dataset_num_proc": 16,
217
+ "load_from_cache_file": false,
218
+ "dataset_shuffle": true,
219
+ "val_dataset_shuffle": false,
220
+ "streaming": false,
221
+ "interleave_prob": null,
222
+ "stopping_strategy": "first_exhausted",
223
+ "shuffle_buffer_size": 1000,
224
+ "download_mode": "reuse_dataset_if_exists",
225
+ "columns": {},
226
+ "strict": false,
227
+ "model_name": null,
228
+ "model_author": null,
229
+ "custom_dataset_info": [
230
+ "swift/llm/dataset/data/custom_dataset_info_qwen35.json"
231
+ ],
232
+ "quant_method": null,
233
+ "quant_bits": null,
234
+ "hqq_axis": null,
235
+ "bnb_4bit_compute_dtype": "bfloat16",
236
+ "bnb_4bit_quant_type": "nf4",
237
+ "bnb_4bit_use_double_quant": true,
238
+ "bnb_4bit_quant_storage": null,
239
+ "max_new_tokens": 2048,
240
+ "temperature": 0.0,
241
+ "top_k": null,
242
+ "top_p": null,
243
+ "repetition_penalty": null,
244
+ "num_beams": 1,
245
+ "stream": false,
246
+ "stop_words": [],
247
+ "logprobs": false,
248
+ "top_logprobs": null,
249
+ "structured_outputs_regex": null,
250
+ "ckpt_dir": null,
251
+ "lora_modules": [],
252
+ "train_type": "lora",
253
+ "adapters": [],
254
+ "external_plugins": [],
255
+ "model_kwargs": {},
256
+ "load_args": false,
257
+ "load_data_args": false,
258
+ "packing": false,
259
+ "packing_length": null,
260
+ "packing_num_proc": 1,
261
+ "lazy_tokenize": true,
262
+ "custom_register_path": [],
263
+ "use_hf": false,
264
+ "ignore_args_error": false,
265
+ "use_swift_lora": false,
266
+ "freeze_parameters": [],
267
+ "freeze_parameters_regex": null,
268
+ "freeze_parameters_ratio": 0.0,
269
+ "trainable_parameters": [],
270
+ "trainable_parameters_regex": null,
271
+ "freeze_llm": false,
272
+ "freeze_vit": false,
273
+ "freeze_aligner": true,
274
+ "target_modules": [
275
+ "all-linear"
276
+ ],
277
+ "target_regex": null,
278
+ "target_parameters": null,
279
+ "modules_to_save": [],
280
+ "lora_rank": 8,
281
+ "lora_alpha": 32,
282
+ "lora_dropout": 0.05,
283
+ "lora_bias": "none",
284
+ "lora_dtype": null,
285
+ "lorap_lr_ratio": null,
286
+ "use_rslora": false,
287
+ "use_dora": false,
288
+ "lora_ga_batch_size": 2,
289
+ "lora_ga_iters": 2,
290
+ "lora_ga_max_length": 1024,
291
+ "lora_ga_direction": "ArB2r",
292
+ "lora_ga_scale": "stable",
293
+ "lora_ga_stable_gamma": 16,
294
+ "init_weights": true,
295
+ "fourier_n_frequency": 2000,
296
+ "fourier_scaling": 300.0,
297
+ "boft_block_size": 4,
298
+ "boft_block_num": 0,
299
+ "boft_n_butterfly_factor": 1,
300
+ "boft_dropout": 0.0,
301
+ "vera_rank": 256,
302
+ "vera_projection_prng_key": 0,
303
+ "vera_dropout": 0.0,
304
+ "vera_d_initial": 0.1,
305
+ "adapter_act": "gelu",
306
+ "adapter_length": 128,
307
+ "use_galore": false,
308
+ "galore_target_modules": null,
309
+ "galore_rank": 128,
310
+ "galore_update_proj_gap": 50,
311
+ "galore_scale": 1.0,
312
+ "galore_proj_type": "std",
313
+ "galore_optim_per_parameter": false,
314
+ "galore_with_embedding": false,
315
+ "galore_quantization": false,
316
+ "galore_proj_quant": false,
317
+ "galore_proj_bits": 4,
318
+ "galore_proj_group_size": 256,
319
+ "galore_cos_threshold": 0.4,
320
+ "galore_gamma_proj": 2,
321
+ "galore_queue_size": 5,
322
+ "adalora_target_r": 8,
323
+ "adalora_init_r": 12,
324
+ "adalora_tinit": 0,
325
+ "adalora_tfinal": 0,
326
+ "adalora_deltaT": 1,
327
+ "adalora_beta1": 0.85,
328
+ "adalora_beta2": 0.85,
329
+ "adalora_orth_reg_weight": 0.5,
330
+ "llamapro_num_new_blocks": 4,
331
+ "llamapro_num_groups": null,
332
+ "lisa_activated_layers": 0,
333
+ "lisa_step_interval": 20,
334
+ "reft_layer_key": null,
335
+ "reft_layers": null,
336
+ "reft_rank": 4,
337
+ "reft_intervention_type": "LoreftIntervention",
338
+ "reft_args": null,
339
+ "swanlab_token": null,
340
+ "swanlab_project": "ms-swift",
341
+ "swanlab_workspace": null,
342
+ "swanlab_exp_name": null,
343
+ "swanlab_notification_method": null,
344
+ "swanlab_webhook_url": null,
345
+ "swanlab_secret": null,
346
+ "swanlab_mode": "cloud",
347
+ "add_version": true,
348
+ "create_checkpoint_symlink": false,
349
+ "zero_hpz_partition_size": null,
350
+ "deepspeed_autotp_size": null,
351
+ "early_stop_interval": null,
352
+ "rank": 0,
353
+ "global_world_size": 2,
354
+ "local_world_size": 2,
355
+ "model_suffix": "Qwen3-VL-2B-Instruct",
356
+ "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='<release-local-path-redacted>, torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)",
357
+ "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=<function get_model_tokenizer_qwen3_vl at 0x74d0706d8790>, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])",
358
+ "model_dir": "<release-local-path-redacted>",
359
+ "_val_dataset_exists": [
360
+ "KEPENGXU/TEST_RAW_FULL"
361
+ ],
362
+ "hub": "<class 'swift.hub.hub.MSHub'>",
363
+ "evaluation_strategy": "steps",
364
+ "training_args": "Seq2SeqTrainingArguments(output_dir='<release-local-path-redacted>, overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_strategy=<IntervalStrategy.STEPS: 'steps'>, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=2, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=0.0001, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=3.0, max_steps=100000, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, lr_scheduler_kwargs=None, warmup_ratio=0.05, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='<release-local-path-redacted>, logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=5000, save_total_limit=20, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=5000, dataloader_num_workers=8, dataloader_prefetch_factor=2, past_index=-1, run_name='<release-local-path-redacted>, disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed=None, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH_FUSED: 'adamw_torch_fused'>, optim_args=None, adafactor=False, group_by_length=False, group_by_size=True, length_column_name='length', report_to=['tensorboard'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint='<release-local-path-redacted>, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='lora', local_repo_path=None, galore_config=None, task_type='causal_lm', problem_type=None)",
365
+ "_config_path": "<release-local-path-redacted>"
366
+ }
output-Qwen3-VL-2B-Instruct/v8-20260421-133546/checkpoint-95000/deploy_result/20260526-234816.jsonl ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ {"response": "The vertical pole visible through the windshield is **yellow**.\n\nIt appears to be a safety or marker pole, commonly seen on roads or in construction zones, and stands out against the snowy background outside.", "infer_request": {"messages": [{"role": "user", "content": [{"type": "image_url", "image_url": {"url": "<<<base64:data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAoAA..>>>"}}, {"type": "text", "text": "This is a linear Image with Metadata: ISO: 250, Exposure Time: 1/640, Aperture: f/9. What is the color of the vertical pole visible through the windshield?"}]}], "images": [], "audios": [], "videos": [], "tools": null, "objects": {}}, "generation_config": "GenerationConfig({'do_sample': True, 'num_assistant_tokens': None, 'num_assistant_tokens_schedule': None, 'return_dict_in_generate': True, 'stop_words': ['<|endoftext|>', '<|im_end|>'], 'temperature': 0.7, 'top_k': 20, 'top_p': 0.8})"}
2
+ {"response": "The vertical pole visible through the windshield is **yellow**.\n\nIt appears to be part of a safety or warning system mounted on the vehicle’s roof or exterior structure, and its bright yellow color stands out against the snowy background and darker interior of the truck.", "infer_request": {"messages": [{"role": "user", "content": [{"type": "image_url", "image_url": {"url": "<<<base64:data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAoAA..>>>"}}, {"type": "text", "text": "This is a linear Image with Metadata: ISO: 250, Exposure Time: 1/640, Aperture: f/9. What is the color of the vertical pole visible through the windshield?"}]}], "images": [], "audios": [], "videos": [], "tools": null, "objects": {}}, "generation_config": "GenerationConfig({'do_sample': True, 'num_assistant_tokens': None, 'num_assistant_tokens_schedule': None, 'return_dict_in_generate': True, 'stop_words': ['<|endoftext|>', '<|im_end|>'], 'temperature': 0.7, 'top_k': 20, 'top_p': 0.8})"}
output-Qwen3-VL-4B-Instruct/v12-20260425-113029/checkpoint-85000/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen3-VL-4B-Instruct
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:Qwen/Qwen3-VL-4B-Instruct
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.17.1
output-Qwen3-VL-4B-Instruct/v12-20260425-113029/checkpoint-85000/adapter_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "Qwen/Qwen3-VL-4B-Instruct",
5
+ "bias": "none",
6
+ "corda_config": null,
7
+ "eva_config": null,
8
+ "exclude_modules": null,
9
+ "fan_in_fan_out": false,
10
+ "inference_mode": true,
11
+ "init_lora_weights": true,
12
+ "layer_replication": null,
13
+ "layers_pattern": null,
14
+ "layers_to_transform": null,
15
+ "loftq_config": {},
16
+ "lora_alpha": 32,
17
+ "lora_bias": false,
18
+ "lora_dropout": 0.05,
19
+ "megatron_config": null,
20
+ "megatron_core": "megatron.core",
21
+ "modules_to_save": [],
22
+ "peft_type": "LORA",
23
+ "qalora_group_size": 16,
24
+ "r": 8,
25
+ "rank_pattern": {},
26
+ "revision": null,
27
+ "target_modules": "^(model.language_model.*\\.(gate_proj|v_proj|k_proj|o_proj|down_proj|up_proj|q_proj)|(?!(model.visual.merger|model.visual.deepstack_merger_list))model.visual.*\\.(attn.proj|qkv|camera_metadata_proj.0|camera_metadata_proj.2|linear_fc1|linear_fc2))$",
28
+ "target_parameters": null,
29
+ "task_type": "CAUSAL_LM",
30
+ "trainable_token_indices": null,
31
+ "use_dora": false,
32
+ "use_qalora": false,
33
+ "use_rslora": false
34
+ }
output-Qwen3-VL-4B-Instruct/v12-20260425-113029/checkpoint-85000/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59a402a46e0b3724c1c5aa52fc056d44e0106dbd53abad08e6e793d02f809dcf
3
+ size 78842264
output-Qwen3-VL-4B-Instruct/v12-20260425-113029/checkpoint-85000/additional_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
output-Qwen3-VL-4B-Instruct/v12-20260425-113029/checkpoint-85000/args.json ADDED
@@ -0,0 +1,366 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "<release-local-path-redacted>",
3
+ "overwrite_output_dir": false,
4
+ "do_train": false,
5
+ "do_eval": false,
6
+ "do_predict": false,
7
+ "eval_strategy": "steps",
8
+ "prediction_loss_only": false,
9
+ "per_device_train_batch_size": 1,
10
+ "per_device_eval_batch_size": 2,
11
+ "per_gpu_train_batch_size": null,
12
+ "per_gpu_eval_batch_size": null,
13
+ "gradient_accumulation_steps": 4,
14
+ "eval_accumulation_steps": null,
15
+ "eval_delay": 0,
16
+ "torch_empty_cache_steps": null,
17
+ "learning_rate": 0.0001,
18
+ "weight_decay": 0.1,
19
+ "adam_beta1": 0.9,
20
+ "adam_beta2": 0.95,
21
+ "adam_epsilon": 1e-08,
22
+ "max_grad_norm": 1.0,
23
+ "num_train_epochs": 3.0,
24
+ "max_steps": 100000,
25
+ "lr_scheduler_type": "cosine",
26
+ "lr_scheduler_kwargs": null,
27
+ "warmup_ratio": 0.05,
28
+ "warmup_steps": 0,
29
+ "log_level": "passive",
30
+ "log_level_replica": "warning",
31
+ "log_on_each_node": true,
32
+ "logging_dir": "<release-local-path-redacted>",
33
+ "logging_strategy": "steps",
34
+ "logging_first_step": true,
35
+ "logging_steps": 5,
36
+ "logging_nan_inf_filter": true,
37
+ "save_strategy": "steps",
38
+ "save_steps": 5000.0,
39
+ "save_total_limit": 20,
40
+ "save_safetensors": true,
41
+ "save_on_each_node": false,
42
+ "save_only_model": false,
43
+ "restore_callback_states_from_checkpoint": false,
44
+ "no_cuda": false,
45
+ "use_cpu": false,
46
+ "use_mps_device": false,
47
+ "seed": 42,
48
+ "data_seed": 42,
49
+ "jit_mode_eval": false,
50
+ "bf16": true,
51
+ "fp16": false,
52
+ "fp16_opt_level": "O1",
53
+ "half_precision_backend": "auto",
54
+ "bf16_full_eval": false,
55
+ "fp16_full_eval": false,
56
+ "tf32": null,
57
+ "local_rank": 0,
58
+ "ddp_backend": null,
59
+ "tpu_num_cores": null,
60
+ "tpu_metrics_debug": false,
61
+ "debug": null,
62
+ "dataloader_drop_last": true,
63
+ "eval_steps": 5000.0,
64
+ "dataloader_num_workers": 8,
65
+ "dataloader_prefetch_factor": null,
66
+ "past_index": -1,
67
+ "run_name": "<release-local-path-redacted>",
68
+ "disable_tqdm": null,
69
+ "remove_unused_columns": false,
70
+ "label_names": null,
71
+ "load_best_model_at_end": false,
72
+ "metric_for_best_model": "loss",
73
+ "greater_is_better": false,
74
+ "ignore_data_skip": false,
75
+ "fsdp": [],
76
+ "fsdp_min_num_params": 0,
77
+ "fsdp_config": null,
78
+ "fsdp_transformer_layer_cls_to_wrap": null,
79
+ "accelerator_config": {
80
+ "dispatch_batches": false
81
+ },
82
+ "parallelism_config": null,
83
+ "deepspeed": null,
84
+ "label_smoothing_factor": 0.0,
85
+ "optim": "adamw_torch_fused",
86
+ "optim_args": null,
87
+ "adafactor": false,
88
+ "group_by_length": false,
89
+ "group_by_size": true,
90
+ "length_column_name": "length",
91
+ "report_to": [
92
+ "tensorboard"
93
+ ],
94
+ "project": "huggingface",
95
+ "trackio_space_id": "trackio",
96
+ "ddp_find_unused_parameters": null,
97
+ "ddp_bucket_cap_mb": null,
98
+ "ddp_broadcast_buffers": null,
99
+ "dataloader_pin_memory": true,
100
+ "dataloader_persistent_workers": false,
101
+ "skip_memory_metrics": true,
102
+ "use_legacy_prediction_loop": false,
103
+ "push_to_hub": false,
104
+ "resume_from_checkpoint": "<release-local-path-redacted>",
105
+ "hub_model_id": null,
106
+ "hub_strategy": "every_save",
107
+ "hub_token": null,
108
+ "hub_private_repo": null,
109
+ "hub_always_push": false,
110
+ "hub_revision": null,
111
+ "gradient_checkpointing": true,
112
+ "gradient_checkpointing_kwargs": null,
113
+ "include_inputs_for_metrics": false,
114
+ "include_for_metrics": [],
115
+ "eval_do_concat_batches": true,
116
+ "fp16_backend": "auto",
117
+ "push_to_hub_model_id": null,
118
+ "push_to_hub_organization": null,
119
+ "push_to_hub_token": null,
120
+ "mp_parameters": "",
121
+ "auto_find_batch_size": false,
122
+ "full_determinism": false,
123
+ "torchdynamo": null,
124
+ "ray_scope": "last",
125
+ "ddp_timeout": 18000000,
126
+ "torch_compile": false,
127
+ "torch_compile_backend": null,
128
+ "torch_compile_mode": null,
129
+ "include_tokens_per_second": false,
130
+ "include_num_input_tokens_seen": false,
131
+ "neftune_noise_alpha": null,
132
+ "optim_target_modules": null,
133
+ "batch_eval_metrics": false,
134
+ "eval_on_start": false,
135
+ "use_liger_kernel": false,
136
+ "liger_kernel_config": null,
137
+ "eval_use_gather_object": false,
138
+ "average_tokens_across_devices": true,
139
+ "sortish_sampler": false,
140
+ "predict_with_generate": false,
141
+ "generation_max_length": null,
142
+ "generation_num_beams": null,
143
+ "generation_config": null,
144
+ "tuner_backend": "peft",
145
+ "vit_gradient_checkpointing": false,
146
+ "router_aux_loss_coef": 0.0,
147
+ "enable_dft_loss": false,
148
+ "enable_channel_loss": false,
149
+ "check_model": true,
150
+ "acc_strategy": "token",
151
+ "train_dataloader_shuffle": true,
152
+ "max_epochs": null,
153
+ "aligner_lr": null,
154
+ "vit_lr": null,
155
+ "use_logits_to_keep": null,
156
+ "ds3_gather_for_generation": true,
157
+ "resume_only_model": false,
158
+ "optimizer": null,
159
+ "loss_type": null,
160
+ "metric": null,
161
+ "eval_use_evalscope": false,
162
+ "eval_dataset": [],
163
+ "eval_dataset_args": null,
164
+ "eval_limit": null,
165
+ "eval_generation_config": null,
166
+ "extra_eval_args": null,
167
+ "use_flash_ckpt": false,
168
+ "use_ray": false,
169
+ "ray_exp_name": null,
170
+ "device_groups": null,
171
+ "model": "Qwen/Qwen3-VL-4B-Instruct",
172
+ "model_type": "qwen3_vl",
173
+ "model_revision": null,
174
+ "task_type": "causal_lm",
175
+ "torch_dtype": "bfloat16",
176
+ "attn_impl": "flash_attn",
177
+ "new_special_tokens": [],
178
+ "num_labels": null,
179
+ "problem_type": null,
180
+ "rope_scaling": null,
181
+ "device_map": null,
182
+ "max_memory": {},
183
+ "max_model_len": null,
184
+ "local_repo_path": null,
185
+ "init_strategy": null,
186
+ "camera_metadata_residual_enabled": true,
187
+ "camera_metadata_residual_num_layers": 2,
188
+ "camera_metadata_require_all": false,
189
+ "camera_metadata_norm_scheme": "photographic_log",
190
+ "template": "qwen3_vl",
191
+ "system": null,
192
+ "max_length": 5120,
193
+ "truncation_strategy": "delete",
194
+ "max_pixels": null,
195
+ "agent_template": null,
196
+ "norm_bbox": null,
197
+ "use_chat_template": true,
198
+ "padding_side": "right",
199
+ "padding_free": false,
200
+ "loss_scale": "default",
201
+ "sequence_parallel_size": 1,
202
+ "template_backend": "swift",
203
+ "response_prefix": null,
204
+ "enable_thinking": null,
205
+ "add_non_thinking_prefix": true,
206
+ "dataset": [
207
+ "KEPENGXU/TRAIN_BALANCE_150K_META",
208
+ "KEPENGXU/TRAIN_BALANCE_150K_META_EXTRA"
209
+ ],
210
+ "val_dataset": [
211
+ "KEPENGXU/TEST_RAW_FULL"
212
+ ],
213
+ "cached_dataset": [],
214
+ "cached_val_dataset": [],
215
+ "split_dataset_ratio": 0.0,
216
+ "dataset_num_proc": 16,
217
+ "load_from_cache_file": false,
218
+ "dataset_shuffle": true,
219
+ "val_dataset_shuffle": false,
220
+ "streaming": false,
221
+ "interleave_prob": null,
222
+ "stopping_strategy": "first_exhausted",
223
+ "shuffle_buffer_size": 1000,
224
+ "download_mode": "reuse_dataset_if_exists",
225
+ "columns": {},
226
+ "strict": false,
227
+ "model_name": null,
228
+ "model_author": null,
229
+ "custom_dataset_info": [
230
+ "swift/llm/dataset/data/custom_dataset_info_qwen35.json"
231
+ ],
232
+ "quant_method": null,
233
+ "quant_bits": null,
234
+ "hqq_axis": null,
235
+ "bnb_4bit_compute_dtype": "bfloat16",
236
+ "bnb_4bit_quant_type": "nf4",
237
+ "bnb_4bit_use_double_quant": true,
238
+ "bnb_4bit_quant_storage": null,
239
+ "max_new_tokens": 2048,
240
+ "temperature": 0.0,
241
+ "top_k": null,
242
+ "top_p": null,
243
+ "repetition_penalty": null,
244
+ "num_beams": 1,
245
+ "stream": false,
246
+ "stop_words": [],
247
+ "logprobs": false,
248
+ "top_logprobs": null,
249
+ "structured_outputs_regex": null,
250
+ "ckpt_dir": null,
251
+ "lora_modules": [],
252
+ "train_type": "lora",
253
+ "adapters": [],
254
+ "external_plugins": [],
255
+ "model_kwargs": {},
256
+ "load_args": false,
257
+ "load_data_args": false,
258
+ "packing": false,
259
+ "packing_length": null,
260
+ "packing_num_proc": 1,
261
+ "lazy_tokenize": true,
262
+ "custom_register_path": [],
263
+ "use_hf": false,
264
+ "ignore_args_error": false,
265
+ "use_swift_lora": false,
266
+ "freeze_parameters": [],
267
+ "freeze_parameters_regex": null,
268
+ "freeze_parameters_ratio": 0.0,
269
+ "trainable_parameters": [],
270
+ "trainable_parameters_regex": null,
271
+ "freeze_llm": false,
272
+ "freeze_vit": false,
273
+ "freeze_aligner": true,
274
+ "target_modules": [
275
+ "all-linear"
276
+ ],
277
+ "target_regex": null,
278
+ "target_parameters": null,
279
+ "modules_to_save": [],
280
+ "lora_rank": 8,
281
+ "lora_alpha": 32,
282
+ "lora_dropout": 0.05,
283
+ "lora_bias": "none",
284
+ "lora_dtype": null,
285
+ "lorap_lr_ratio": null,
286
+ "use_rslora": false,
287
+ "use_dora": false,
288
+ "lora_ga_batch_size": 2,
289
+ "lora_ga_iters": 2,
290
+ "lora_ga_max_length": 1024,
291
+ "lora_ga_direction": "ArB2r",
292
+ "lora_ga_scale": "stable",
293
+ "lora_ga_stable_gamma": 16,
294
+ "init_weights": true,
295
+ "fourier_n_frequency": 2000,
296
+ "fourier_scaling": 300.0,
297
+ "boft_block_size": 4,
298
+ "boft_block_num": 0,
299
+ "boft_n_butterfly_factor": 1,
300
+ "boft_dropout": 0.0,
301
+ "vera_rank": 256,
302
+ "vera_projection_prng_key": 0,
303
+ "vera_dropout": 0.0,
304
+ "vera_d_initial": 0.1,
305
+ "adapter_act": "gelu",
306
+ "adapter_length": 128,
307
+ "use_galore": false,
308
+ "galore_target_modules": null,
309
+ "galore_rank": 128,
310
+ "galore_update_proj_gap": 50,
311
+ "galore_scale": 1.0,
312
+ "galore_proj_type": "std",
313
+ "galore_optim_per_parameter": false,
314
+ "galore_with_embedding": false,
315
+ "galore_quantization": false,
316
+ "galore_proj_quant": false,
317
+ "galore_proj_bits": 4,
318
+ "galore_proj_group_size": 256,
319
+ "galore_cos_threshold": 0.4,
320
+ "galore_gamma_proj": 2,
321
+ "galore_queue_size": 5,
322
+ "adalora_target_r": 8,
323
+ "adalora_init_r": 12,
324
+ "adalora_tinit": 0,
325
+ "adalora_tfinal": 0,
326
+ "adalora_deltaT": 1,
327
+ "adalora_beta1": 0.85,
328
+ "adalora_beta2": 0.85,
329
+ "adalora_orth_reg_weight": 0.5,
330
+ "llamapro_num_new_blocks": 4,
331
+ "llamapro_num_groups": null,
332
+ "lisa_activated_layers": 0,
333
+ "lisa_step_interval": 20,
334
+ "reft_layer_key": null,
335
+ "reft_layers": null,
336
+ "reft_rank": 4,
337
+ "reft_intervention_type": "LoreftIntervention",
338
+ "reft_args": null,
339
+ "swanlab_token": null,
340
+ "swanlab_project": "ms-swift",
341
+ "swanlab_workspace": null,
342
+ "swanlab_exp_name": null,
343
+ "swanlab_notification_method": null,
344
+ "swanlab_webhook_url": null,
345
+ "swanlab_secret": null,
346
+ "swanlab_mode": "cloud",
347
+ "add_version": true,
348
+ "create_checkpoint_symlink": false,
349
+ "zero_hpz_partition_size": null,
350
+ "deepspeed_autotp_size": null,
351
+ "early_stop_interval": null,
352
+ "rank": 0,
353
+ "global_world_size": 2,
354
+ "local_world_size": 2,
355
+ "model_suffix": "Qwen3-VL-4B-Instruct",
356
+ "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='<release-local-path-redacted>, torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)",
357
+ "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=<function get_model_tokenizer_qwen3_vl at 0x7031b788c790>, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])",
358
+ "model_dir": "<release-local-path-redacted>",
359
+ "_val_dataset_exists": [
360
+ "KEPENGXU/TEST_RAW_FULL"
361
+ ],
362
+ "hub": "<class 'swift.hub.hub.MSHub'>",
363
+ "evaluation_strategy": "steps",
364
+ "training_args": "Seq2SeqTrainingArguments(output_dir='<release-local-path-redacted>, overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_strategy=<IntervalStrategy.STEPS: 'steps'>, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=2, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=0.0001, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=3.0, max_steps=100000, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, lr_scheduler_kwargs=None, warmup_ratio=0.05, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='<release-local-path-redacted>, logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=5000, save_total_limit=20, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=True, eval_steps=5000, dataloader_num_workers=8, dataloader_prefetch_factor=2, past_index=-1, run_name='<release-local-path-redacted>, disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed=None, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH_FUSED: 'adamw_torch_fused'>, optim_args=None, adafactor=False, group_by_length=False, group_by_size=True, length_column_name='length', report_to=['tensorboard'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint='<release-local-path-redacted>, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='lora', local_repo_path=None, galore_config=None, task_type='causal_lm', problem_type=None)",
365
+ "_config_path": "<release-local-path-redacted>"
366
+ }
output-Qwen3-VL-8B-Instruct/v2-20260423-205317/checkpoint-95000/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen3-VL-8B-Instruct
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:Qwen/Qwen3-VL-8B-Instruct
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.17.1
output-Qwen3-VL-8B-Instruct/v2-20260423-205317/checkpoint-95000/adapter_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "Qwen/Qwen3-VL-8B-Instruct",
5
+ "bias": "none",
6
+ "corda_config": null,
7
+ "eva_config": null,
8
+ "exclude_modules": null,
9
+ "fan_in_fan_out": false,
10
+ "inference_mode": true,
11
+ "init_lora_weights": true,
12
+ "layer_replication": null,
13
+ "layers_pattern": null,
14
+ "layers_to_transform": null,
15
+ "loftq_config": {},
16
+ "lora_alpha": 32,
17
+ "lora_bias": false,
18
+ "lora_dropout": 0.05,
19
+ "megatron_config": null,
20
+ "megatron_core": "megatron.core",
21
+ "modules_to_save": [],
22
+ "peft_type": "LORA",
23
+ "qalora_group_size": 16,
24
+ "r": 8,
25
+ "rank_pattern": {},
26
+ "revision": null,
27
+ "target_modules": "^(model.language_model.*\\.(q_proj|k_proj|up_proj|o_proj|v_proj|down_proj|gate_proj)|(?!(model.visual.merger|model.visual.deepstack_merger_list))model.visual.*\\.(qkv|linear_fc2|linear_fc1|camera_metadata_proj.2|attn.proj|camera_metadata_proj.0))$",
28
+ "target_parameters": null,
29
+ "task_type": "CAUSAL_LM",
30
+ "trainable_token_indices": null,
31
+ "use_dora": false,
32
+ "use_qalora": false,
33
+ "use_rslora": false
34
+ }
output-Qwen3-VL-8B-Instruct/v2-20260423-205317/checkpoint-95000/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea38db41f51bbd8c87f97537911fad08df029fe5bf7f12e50501a19574e2fb36
3
+ size 102908744
output-Qwen3-VL-8B-Instruct/v2-20260423-205317/checkpoint-95000/additional_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
output-Qwen3-VL-8B-Instruct/v2-20260423-205317/checkpoint-95000/args.json ADDED
@@ -0,0 +1,366 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "<release-local-path-redacted>",
3
+ "overwrite_output_dir": false,
4
+ "do_train": false,
5
+ "do_eval": false,
6
+ "do_predict": false,
7
+ "eval_strategy": "steps",
8
+ "prediction_loss_only": false,
9
+ "per_device_train_batch_size": 1,
10
+ "per_device_eval_batch_size": 2,
11
+ "per_gpu_train_batch_size": null,
12
+ "per_gpu_eval_batch_size": null,
13
+ "gradient_accumulation_steps": 4,
14
+ "eval_accumulation_steps": null,
15
+ "eval_delay": 0,
16
+ "torch_empty_cache_steps": null,
17
+ "learning_rate": 0.0001,
18
+ "weight_decay": 0.1,
19
+ "adam_beta1": 0.9,
20
+ "adam_beta2": 0.95,
21
+ "adam_epsilon": 1e-08,
22
+ "max_grad_norm": 1.0,
23
+ "num_train_epochs": 3.0,
24
+ "max_steps": 100000,
25
+ "lr_scheduler_type": "cosine",
26
+ "lr_scheduler_kwargs": null,
27
+ "warmup_ratio": 0.05,
28
+ "warmup_steps": 0,
29
+ "log_level": "passive",
30
+ "log_level_replica": "warning",
31
+ "log_on_each_node": true,
32
+ "logging_dir": "<release-local-path-redacted>",
33
+ "logging_strategy": "steps",
34
+ "logging_first_step": true,
35
+ "logging_steps": 5,
36
+ "logging_nan_inf_filter": true,
37
+ "save_strategy": "steps",
38
+ "save_steps": 5000.0,
39
+ "save_total_limit": 20,
40
+ "save_safetensors": true,
41
+ "save_on_each_node": false,
42
+ "save_only_model": false,
43
+ "restore_callback_states_from_checkpoint": false,
44
+ "no_cuda": false,
45
+ "use_cpu": false,
46
+ "use_mps_device": false,
47
+ "seed": 42,
48
+ "data_seed": 42,
49
+ "jit_mode_eval": false,
50
+ "bf16": true,
51
+ "fp16": false,
52
+ "fp16_opt_level": "O1",
53
+ "half_precision_backend": "auto",
54
+ "bf16_full_eval": false,
55
+ "fp16_full_eval": false,
56
+ "tf32": null,
57
+ "local_rank": 0,
58
+ "ddp_backend": null,
59
+ "tpu_num_cores": null,
60
+ "tpu_metrics_debug": false,
61
+ "debug": null,
62
+ "dataloader_drop_last": true,
63
+ "eval_steps": 5000.0,
64
+ "dataloader_num_workers": 8,
65
+ "dataloader_prefetch_factor": null,
66
+ "past_index": -1,
67
+ "run_name": "<release-local-path-redacted>",
68
+ "disable_tqdm": null,
69
+ "remove_unused_columns": false,
70
+ "label_names": null,
71
+ "load_best_model_at_end": false,
72
+ "metric_for_best_model": "loss",
73
+ "greater_is_better": false,
74
+ "ignore_data_skip": false,
75
+ "fsdp": [],
76
+ "fsdp_min_num_params": 0,
77
+ "fsdp_config": null,
78
+ "fsdp_transformer_layer_cls_to_wrap": null,
79
+ "accelerator_config": {
80
+ "dispatch_batches": false
81
+ },
82
+ "parallelism_config": null,
83
+ "deepspeed": null,
84
+ "label_smoothing_factor": 0.0,
85
+ "optim": "adamw_torch_fused",
86
+ "optim_args": null,
87
+ "adafactor": false,
88
+ "group_by_length": false,
89
+ "group_by_size": true,
90
+ "length_column_name": "length",
91
+ "report_to": [
92
+ "tensorboard"
93
+ ],
94
+ "project": "huggingface",
95
+ "trackio_space_id": "trackio",
96
+ "ddp_find_unused_parameters": null,
97
+ "ddp_bucket_cap_mb": null,
98
+ "ddp_broadcast_buffers": null,
99
+ "dataloader_pin_memory": true,
100
+ "dataloader_persistent_workers": false,
101
+ "skip_memory_metrics": true,
102
+ "use_legacy_prediction_loop": false,
103
+ "push_to_hub": false,
104
+ "resume_from_checkpoint": "<release-local-path-redacted>",
105
+ "hub_model_id": null,
106
+ "hub_strategy": "every_save",
107
+ "hub_token": null,
108
+ "hub_private_repo": null,
109
+ "hub_always_push": false,
110
+ "hub_revision": null,
111
+ "gradient_checkpointing": true,
112
+ "gradient_checkpointing_kwargs": null,
113
+ "include_inputs_for_metrics": false,
114
+ "include_for_metrics": [],
115
+ "eval_do_concat_batches": true,
116
+ "fp16_backend": "auto",
117
+ "push_to_hub_model_id": null,
118
+ "push_to_hub_organization": null,
119
+ "push_to_hub_token": null,
120
+ "mp_parameters": "",
121
+ "auto_find_batch_size": false,
122
+ "full_determinism": false,
123
+ "torchdynamo": null,
124
+ "ray_scope": "last",
125
+ "ddp_timeout": 18000000,
126
+ "torch_compile": false,
127
+ "torch_compile_backend": null,
128
+ "torch_compile_mode": null,
129
+ "include_tokens_per_second": false,
130
+ "include_num_input_tokens_seen": false,
131
+ "neftune_noise_alpha": null,
132
+ "optim_target_modules": null,
133
+ "batch_eval_metrics": false,
134
+ "eval_on_start": false,
135
+ "use_liger_kernel": false,
136
+ "liger_kernel_config": null,
137
+ "eval_use_gather_object": false,
138
+ "average_tokens_across_devices": true,
139
+ "sortish_sampler": false,
140
+ "predict_with_generate": false,
141
+ "generation_max_length": null,
142
+ "generation_num_beams": null,
143
+ "generation_config": null,
144
+ "tuner_backend": "peft",
145
+ "vit_gradient_checkpointing": false,
146
+ "router_aux_loss_coef": 0.0,
147
+ "enable_dft_loss": false,
148
+ "enable_channel_loss": false,
149
+ "check_model": true,
150
+ "acc_strategy": "token",
151
+ "train_dataloader_shuffle": true,
152
+ "max_epochs": null,
153
+ "aligner_lr": null,
154
+ "vit_lr": null,
155
+ "use_logits_to_keep": null,
156
+ "ds3_gather_for_generation": true,
157
+ "resume_only_model": false,
158
+ "optimizer": null,
159
+ "loss_type": null,
160
+ "metric": null,
161
+ "eval_use_evalscope": false,
162
+ "eval_dataset": [],
163
+ "eval_dataset_args": null,
164
+ "eval_limit": null,
165
+ "eval_generation_config": null,
166
+ "extra_eval_args": null,
167
+ "use_flash_ckpt": false,
168
+ "use_ray": false,
169
+ "ray_exp_name": null,
170
+ "device_groups": null,
171
+ "model": "Qwen/Qwen3-VL-8B-Instruct",
172
+ "model_type": "qwen3_vl",
173
+ "model_revision": null,
174
+ "task_type": "causal_lm",
175
+ "torch_dtype": "bfloat16",
176
+ "attn_impl": "flash_attn",
177
+ "new_special_tokens": [],
178
+ "num_labels": null,
179
+ "problem_type": null,
180
+ "rope_scaling": null,
181
+ "device_map": null,
182
+ "max_memory": {},
183
+ "max_model_len": null,
184
+ "local_repo_path": null,
185
+ "init_strategy": null,
186
+ "camera_metadata_residual_enabled": true,
187
+ "camera_metadata_residual_num_layers": 2,
188
+ "camera_metadata_require_all": false,
189
+ "camera_metadata_norm_scheme": "photographic_log",
190
+ "template": "qwen3_vl",
191
+ "system": null,
192
+ "max_length": 5120,
193
+ "truncation_strategy": "delete",
194
+ "max_pixels": null,
195
+ "agent_template": null,
196
+ "norm_bbox": null,
197
+ "use_chat_template": true,
198
+ "padding_side": "right",
199
+ "padding_free": false,
200
+ "loss_scale": "default",
201
+ "sequence_parallel_size": 1,
202
+ "template_backend": "swift",
203
+ "response_prefix": null,
204
+ "enable_thinking": null,
205
+ "add_non_thinking_prefix": true,
206
+ "dataset": [
207
+ "KEPENGXU/TRAIN_BALANCE_150K_META",
208
+ "KEPENGXU/TRAIN_BALANCE_150K_META_EXTRA"
209
+ ],
210
+ "val_dataset": [
211
+ "KEPENGXU/TEST_RAW_FULL"
212
+ ],
213
+ "cached_dataset": [],
214
+ "cached_val_dataset": [],
215
+ "split_dataset_ratio": 0.0,
216
+ "dataset_num_proc": 16,
217
+ "load_from_cache_file": false,
218
+ "dataset_shuffle": true,
219
+ "val_dataset_shuffle": false,
220
+ "streaming": false,
221
+ "interleave_prob": null,
222
+ "stopping_strategy": "first_exhausted",
223
+ "shuffle_buffer_size": 1000,
224
+ "download_mode": "reuse_dataset_if_exists",
225
+ "columns": {},
226
+ "strict": false,
227
+ "model_name": null,
228
+ "model_author": null,
229
+ "custom_dataset_info": [
230
+ "swift/llm/dataset/data/custom_dataset_info_qwen35.json"
231
+ ],
232
+ "quant_method": null,
233
+ "quant_bits": null,
234
+ "hqq_axis": null,
235
+ "bnb_4bit_compute_dtype": "bfloat16",
236
+ "bnb_4bit_quant_type": "nf4",
237
+ "bnb_4bit_use_double_quant": true,
238
+ "bnb_4bit_quant_storage": null,
239
+ "max_new_tokens": 2048,
240
+ "temperature": 0.0,
241
+ "top_k": null,
242
+ "top_p": null,
243
+ "repetition_penalty": null,
244
+ "num_beams": 1,
245
+ "stream": false,
246
+ "stop_words": [],
247
+ "logprobs": false,
248
+ "top_logprobs": null,
249
+ "structured_outputs_regex": null,
250
+ "ckpt_dir": null,
251
+ "lora_modules": [],
252
+ "train_type": "lora",
253
+ "adapters": [],
254
+ "external_plugins": [],
255
+ "model_kwargs": {},
256
+ "load_args": false,
257
+ "load_data_args": false,
258
+ "packing": false,
259
+ "packing_length": null,
260
+ "packing_num_proc": 1,
261
+ "lazy_tokenize": true,
262
+ "custom_register_path": [],
263
+ "use_hf": false,
264
+ "ignore_args_error": false,
265
+ "use_swift_lora": false,
266
+ "freeze_parameters": [],
267
+ "freeze_parameters_regex": null,
268
+ "freeze_parameters_ratio": 0.0,
269
+ "trainable_parameters": [],
270
+ "trainable_parameters_regex": null,
271
+ "freeze_llm": false,
272
+ "freeze_vit": false,
273
+ "freeze_aligner": true,
274
+ "target_modules": [
275
+ "all-linear"
276
+ ],
277
+ "target_regex": null,
278
+ "target_parameters": null,
279
+ "modules_to_save": [],
280
+ "lora_rank": 8,
281
+ "lora_alpha": 32,
282
+ "lora_dropout": 0.05,
283
+ "lora_bias": "none",
284
+ "lora_dtype": null,
285
+ "lorap_lr_ratio": null,
286
+ "use_rslora": false,
287
+ "use_dora": false,
288
+ "lora_ga_batch_size": 2,
289
+ "lora_ga_iters": 2,
290
+ "lora_ga_max_length": 1024,
291
+ "lora_ga_direction": "ArB2r",
292
+ "lora_ga_scale": "stable",
293
+ "lora_ga_stable_gamma": 16,
294
+ "init_weights": true,
295
+ "fourier_n_frequency": 2000,
296
+ "fourier_scaling": 300.0,
297
+ "boft_block_size": 4,
298
+ "boft_block_num": 0,
299
+ "boft_n_butterfly_factor": 1,
300
+ "boft_dropout": 0.0,
301
+ "vera_rank": 256,
302
+ "vera_projection_prng_key": 0,
303
+ "vera_dropout": 0.0,
304
+ "vera_d_initial": 0.1,
305
+ "adapter_act": "gelu",
306
+ "adapter_length": 128,
307
+ "use_galore": false,
308
+ "galore_target_modules": null,
309
+ "galore_rank": 128,
310
+ "galore_update_proj_gap": 50,
311
+ "galore_scale": 1.0,
312
+ "galore_proj_type": "std",
313
+ "galore_optim_per_parameter": false,
314
+ "galore_with_embedding": false,
315
+ "galore_quantization": false,
316
+ "galore_proj_quant": false,
317
+ "galore_proj_bits": 4,
318
+ "galore_proj_group_size": 256,
319
+ "galore_cos_threshold": 0.4,
320
+ "galore_gamma_proj": 2,
321
+ "galore_queue_size": 5,
322
+ "adalora_target_r": 8,
323
+ "adalora_init_r": 12,
324
+ "adalora_tinit": 0,
325
+ "adalora_tfinal": 0,
326
+ "adalora_deltaT": 1,
327
+ "adalora_beta1": 0.85,
328
+ "adalora_beta2": 0.85,
329
+ "adalora_orth_reg_weight": 0.5,
330
+ "llamapro_num_new_blocks": 4,
331
+ "llamapro_num_groups": null,
332
+ "lisa_activated_layers": 0,
333
+ "lisa_step_interval": 20,
334
+ "reft_layer_key": null,
335
+ "reft_layers": null,
336
+ "reft_rank": 4,
337
+ "reft_intervention_type": "LoreftIntervention",
338
+ "reft_args": null,
339
+ "swanlab_token": null,
340
+ "swanlab_project": "ms-swift",
341
+ "swanlab_workspace": null,
342
+ "swanlab_exp_name": null,
343
+ "swanlab_notification_method": null,
344
+ "swanlab_webhook_url": null,
345
+ "swanlab_secret": null,
346
+ "swanlab_mode": "cloud",
347
+ "add_version": true,
348
+ "create_checkpoint_symlink": false,
349
+ "zero_hpz_partition_size": null,
350
+ "deepspeed_autotp_size": null,
351
+ "early_stop_interval": null,
352
+ "rank": 0,
353
+ "global_world_size": 2,
354
+ "local_world_size": 2,
355
+ "model_suffix": "Qwen3-VL-8B-Instruct",
356
+ "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='<release-local-path-redacted>, torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)",
357
+ "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=<function get_model_tokenizer_qwen3_vl at 0x7f6229fa0790>, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])",
358
+ "model_dir": "<release-local-path-redacted>",
359
+ "_val_dataset_exists": [
360
+ "KEPENGXU/TEST_RAW_FULL"
361
+ ],
362
+ "hub": "<class 'swift.hub.hub.MSHub'>",
363
+ "evaluation_strategy": "steps",
364
+ "training_args": "Seq2SeqTrainingArguments(output_dir='<release-local-path-redacted>, overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_strategy=<IntervalStrategy.STEPS: 'steps'>, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=2, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=0.0001, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=3.0, max_steps=100000, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, lr_scheduler_kwargs=None, warmup_ratio=0.05, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='<release-local-path-redacted>, logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=5000, save_total_limit=20, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=True, eval_steps=5000, dataloader_num_workers=8, dataloader_prefetch_factor=2, past_index=-1, run_name='<release-local-path-redacted>, disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed=None, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH_FUSED: 'adamw_torch_fused'>, optim_args=None, adafactor=False, group_by_length=False, group_by_size=True, length_column_name='length', report_to=['tensorboard'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint='<release-local-path-redacted>, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='lora', local_repo_path=None, galore_config=None, task_type='causal_lm', problem_type=None)",
365
+ "_config_path": "<release-local-path-redacted>"
366
+ }