Commit ·
1a428bf
0
Parent(s):
Duplicate from YinmingHuang/VA-Judger
Browse filesCo-authored-by: Yinming Huang <YinmingHuang@users.noreply.huggingface.co>
- .gitattributes +36 -0
- LTX-RL-VA-Judger/lora/adapter_config.json +45 -0
- LTX-RL-VA-Judger/lora/adapter_model.safetensors +3 -0
- README.md +3 -0
- VA-Judger/args.json +398 -0
- VA-Judger/chat_template.jinja +122 -0
- VA-Judger/chat_template.json +1 -0
- VA-Judger/config.json +430 -0
- VA-Judger/generation_config.json +12 -0
- VA-Judger/latest +1 -0
- VA-Judger/merges.txt +0 -0
- VA-Judger/model-00001-of-00013.safetensors +3 -0
- VA-Judger/model-00002-of-00013.safetensors +3 -0
- VA-Judger/model-00003-of-00013.safetensors +3 -0
- VA-Judger/model-00004-of-00013.safetensors +3 -0
- VA-Judger/model-00005-of-00013.safetensors +3 -0
- VA-Judger/model-00006-of-00013.safetensors +3 -0
- VA-Judger/model-00007-of-00013.safetensors +3 -0
- VA-Judger/model-00008-of-00013.safetensors +3 -0
- VA-Judger/model-00009-of-00013.safetensors +3 -0
- VA-Judger/model-00010-of-00013.safetensors +3 -0
- VA-Judger/model-00011-of-00013.safetensors +3 -0
- VA-Judger/model-00012-of-00013.safetensors +3 -0
- VA-Judger/model-00013-of-00013.safetensors +3 -0
- VA-Judger/model.safetensors.index.json +0 -0
- VA-Judger/preprocessor_config.json +30 -0
- VA-Judger/processor_config.json +109 -0
- VA-Judger/tokenizer.json +3 -0
- VA-Judger/tokenizer_config.json +316 -0
- VA-Judger/trainer_state.json +1962 -0
- VA-Judger/training_args.bin +3 -0
- VA-Judger/vocab.json +0 -0
- VA-Judger/zero_to_fp32.py +760 -0
.gitattributes
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
| 2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
| 3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
| 4 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
| 5 |
+
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
| 6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
| 7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
| 8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
| 9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
| 10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
| 11 |
+
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
| 12 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
| 13 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
| 14 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
| 15 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
| 16 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
| 17 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
| 18 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
| 19 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
| 20 |
+
*.pickle filter=lfs diff=lfs merge=lfs -text
|
| 21 |
+
*.pkl filter=lfs diff=lfs merge=lfs -text
|
| 22 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
| 23 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
| 24 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
| 25 |
+
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 26 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
| 27 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
| 28 |
+
*.tar filter=lfs diff=lfs merge=lfs -text
|
| 29 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
| 30 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
| 31 |
+
*.wasm filter=lfs diff=lfs merge=lfs -text
|
| 32 |
+
*.xz 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 |
+
VA-Judger/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
LTX-RL-VA-Judger/lora/adapter_config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": null,
|
| 6 |
+
"base_model_name_or_path": null,
|
| 7 |
+
"bias": "none",
|
| 8 |
+
"corda_config": null,
|
| 9 |
+
"ensure_weight_tying": false,
|
| 10 |
+
"eva_config": null,
|
| 11 |
+
"exclude_modules": null,
|
| 12 |
+
"fan_in_fan_out": false,
|
| 13 |
+
"inference_mode": false,
|
| 14 |
+
"init_lora_weights": "gaussian",
|
| 15 |
+
"layer_replication": null,
|
| 16 |
+
"layers_pattern": null,
|
| 17 |
+
"layers_to_transform": null,
|
| 18 |
+
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 64,
|
| 20 |
+
"lora_bias": false,
|
| 21 |
+
"lora_dropout": 0.0,
|
| 22 |
+
"megatron_config": null,
|
| 23 |
+
"megatron_core": "megatron.core",
|
| 24 |
+
"modules_to_save": null,
|
| 25 |
+
"peft_type": "LORA",
|
| 26 |
+
"peft_version": "0.18.1",
|
| 27 |
+
"qalora_group_size": 16,
|
| 28 |
+
"r": 32,
|
| 29 |
+
"rank_pattern": {},
|
| 30 |
+
"revision": null,
|
| 31 |
+
"target_modules": [
|
| 32 |
+
"to_q",
|
| 33 |
+
"to_k",
|
| 34 |
+
"net.2",
|
| 35 |
+
"to_out.0",
|
| 36 |
+
"to_v",
|
| 37 |
+
"proj"
|
| 38 |
+
],
|
| 39 |
+
"target_parameters": null,
|
| 40 |
+
"task_type": null,
|
| 41 |
+
"trainable_token_indices": null,
|
| 42 |
+
"use_dora": false,
|
| 43 |
+
"use_qalora": false,
|
| 44 |
+
"use_rslora": false
|
| 45 |
+
}
|
LTX-RL-VA-Judger/lora/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5defa0873a16b87ec2512266e98ad25862623ddcc85fdb21bda0bb815cf01b3
|
| 3 |
+
size 1233512728
|
README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
VA-Judger/args.json
ADDED
|
@@ -0,0 +1,398 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"output_dir": "/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run",
|
| 3 |
+
"per_device_train_batch_size": 4,
|
| 4 |
+
"num_train_epochs": 3.0,
|
| 5 |
+
"max_steps": 2000,
|
| 6 |
+
"learning_rate": 5e-06,
|
| 7 |
+
"lr_scheduler_type": "cosine",
|
| 8 |
+
"lr_scheduler_kwargs": null,
|
| 9 |
+
"warmup_steps": 0,
|
| 10 |
+
"optim": "adamw_torch_fused",
|
| 11 |
+
"optim_args": null,
|
| 12 |
+
"weight_decay": 0.1,
|
| 13 |
+
"adam_beta1": 0.9,
|
| 14 |
+
"adam_beta2": 0.95,
|
| 15 |
+
"adam_epsilon": 1e-08,
|
| 16 |
+
"optim_target_modules": null,
|
| 17 |
+
"gradient_accumulation_steps": 4,
|
| 18 |
+
"average_tokens_across_devices": true,
|
| 19 |
+
"max_grad_norm": 1.0,
|
| 20 |
+
"label_smoothing_factor": 0.0,
|
| 21 |
+
"bf16": true,
|
| 22 |
+
"fp16": false,
|
| 23 |
+
"bf16_full_eval": false,
|
| 24 |
+
"fp16_full_eval": false,
|
| 25 |
+
"tf32": null,
|
| 26 |
+
"gradient_checkpointing": true,
|
| 27 |
+
"gradient_checkpointing_kwargs": null,
|
| 28 |
+
"torch_compile": false,
|
| 29 |
+
"torch_compile_backend": null,
|
| 30 |
+
"torch_compile_mode": null,
|
| 31 |
+
"use_liger_kernel": false,
|
| 32 |
+
"liger_kernel_config": null,
|
| 33 |
+
"use_cache": false,
|
| 34 |
+
"neftune_noise_alpha": null,
|
| 35 |
+
"torch_empty_cache_steps": null,
|
| 36 |
+
"auto_find_batch_size": false,
|
| 37 |
+
"logging_strategy": "steps",
|
| 38 |
+
"logging_steps": 5,
|
| 39 |
+
"logging_first_step": true,
|
| 40 |
+
"log_on_each_node": true,
|
| 41 |
+
"logging_nan_inf_filter": true,
|
| 42 |
+
"include_num_input_tokens_seen": false,
|
| 43 |
+
"log_level": "passive",
|
| 44 |
+
"log_level_replica": "warning",
|
| 45 |
+
"disable_tqdm": null,
|
| 46 |
+
"report_to": [
|
| 47 |
+
"tensorboard"
|
| 48 |
+
],
|
| 49 |
+
"run_name": "/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run",
|
| 50 |
+
"project": "huggingface",
|
| 51 |
+
"trackio_space_id": null,
|
| 52 |
+
"trackio_bucket_id": null,
|
| 53 |
+
"trackio_static_space_id": null,
|
| 54 |
+
"eval_strategy": "no",
|
| 55 |
+
"eval_steps": 100.0,
|
| 56 |
+
"eval_delay": 0,
|
| 57 |
+
"per_device_eval_batch_size": 1,
|
| 58 |
+
"prediction_loss_only": false,
|
| 59 |
+
"eval_on_start": false,
|
| 60 |
+
"eval_do_concat_batches": true,
|
| 61 |
+
"eval_use_gather_object": false,
|
| 62 |
+
"eval_accumulation_steps": null,
|
| 63 |
+
"include_for_metrics": [],
|
| 64 |
+
"batch_eval_metrics": false,
|
| 65 |
+
"save_only_model": false,
|
| 66 |
+
"save_strategy": "steps",
|
| 67 |
+
"save_steps": 100.0,
|
| 68 |
+
"save_on_each_node": false,
|
| 69 |
+
"save_total_limit": 30,
|
| 70 |
+
"enable_jit_checkpoint": false,
|
| 71 |
+
"push_to_hub": false,
|
| 72 |
+
"hub_token": null,
|
| 73 |
+
"hub_private_repo": null,
|
| 74 |
+
"hub_model_id": null,
|
| 75 |
+
"hub_strategy": "every_save",
|
| 76 |
+
"hub_always_push": false,
|
| 77 |
+
"hub_revision": null,
|
| 78 |
+
"load_best_model_at_end": false,
|
| 79 |
+
"metric_for_best_model": "loss",
|
| 80 |
+
"greater_is_better": false,
|
| 81 |
+
"ignore_data_skip": false,
|
| 82 |
+
"restore_callback_states_from_checkpoint": false,
|
| 83 |
+
"full_determinism": false,
|
| 84 |
+
"seed": 42,
|
| 85 |
+
"data_seed": 42,
|
| 86 |
+
"use_cpu": false,
|
| 87 |
+
"accelerator_config": {
|
| 88 |
+
"dispatch_batches": false
|
| 89 |
+
},
|
| 90 |
+
"parallelism_config": null,
|
| 91 |
+
"dataloader_drop_last": false,
|
| 92 |
+
"dataloader_num_workers": 4,
|
| 93 |
+
"dataloader_pin_memory": true,
|
| 94 |
+
"dataloader_persistent_workers": false,
|
| 95 |
+
"dataloader_prefetch_factor": null,
|
| 96 |
+
"remove_unused_columns": true,
|
| 97 |
+
"label_names": null,
|
| 98 |
+
"train_sampling_strategy": "random",
|
| 99 |
+
"length_column_name": "length",
|
| 100 |
+
"ddp_find_unused_parameters": null,
|
| 101 |
+
"ddp_bucket_cap_mb": null,
|
| 102 |
+
"ddp_broadcast_buffers": null,
|
| 103 |
+
"ddp_static_graph": null,
|
| 104 |
+
"ddp_backend": null,
|
| 105 |
+
"ddp_timeout": 18000000,
|
| 106 |
+
"fsdp": [],
|
| 107 |
+
"fsdp_config": null,
|
| 108 |
+
"deepspeed": {
|
| 109 |
+
"fp16": {
|
| 110 |
+
"enabled": "auto",
|
| 111 |
+
"loss_scale": 0,
|
| 112 |
+
"loss_scale_window": 1000,
|
| 113 |
+
"initial_scale_power": 16,
|
| 114 |
+
"hysteresis": 2,
|
| 115 |
+
"min_loss_scale": 1
|
| 116 |
+
},
|
| 117 |
+
"bf16": {
|
| 118 |
+
"enabled": "auto"
|
| 119 |
+
},
|
| 120 |
+
"zero_optimization": {
|
| 121 |
+
"stage": 3,
|
| 122 |
+
"offload_optimizer": {
|
| 123 |
+
"device": "cpu",
|
| 124 |
+
"pin_memory": true
|
| 125 |
+
},
|
| 126 |
+
"offload_param": {
|
| 127 |
+
"device": "cpu",
|
| 128 |
+
"pin_memory": true
|
| 129 |
+
},
|
| 130 |
+
"overlap_comm": false,
|
| 131 |
+
"contiguous_gradients": true,
|
| 132 |
+
"sub_group_size": 1000000000.0,
|
| 133 |
+
"reduce_bucket_size": "auto",
|
| 134 |
+
"stage3_prefetch_bucket_size": "auto",
|
| 135 |
+
"stage3_param_persistence_threshold": "auto",
|
| 136 |
+
"stage3_max_live_parameters": 1000000000.0,
|
| 137 |
+
"stage3_max_reuse_distance": 1000000000.0,
|
| 138 |
+
"stage3_gather_16bit_weights_on_model_save": true
|
| 139 |
+
},
|
| 140 |
+
"gradient_accumulation_steps": "auto",
|
| 141 |
+
"gradient_clipping": "auto",
|
| 142 |
+
"steps_per_print": 2000,
|
| 143 |
+
"train_batch_size": "auto",
|
| 144 |
+
"train_micro_batch_size_per_gpu": "auto",
|
| 145 |
+
"wall_clock_breakdown": false
|
| 146 |
+
},
|
| 147 |
+
"debug": null,
|
| 148 |
+
"skip_memory_metrics": true,
|
| 149 |
+
"do_train": false,
|
| 150 |
+
"do_eval": false,
|
| 151 |
+
"do_predict": false,
|
| 152 |
+
"resume_from_checkpoint": "/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run/checkpoint-1100",
|
| 153 |
+
"warmup_ratio": 0.05,
|
| 154 |
+
"logging_dir": "/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run/runs",
|
| 155 |
+
"local_rank": 0,
|
| 156 |
+
"sortish_sampler": false,
|
| 157 |
+
"predict_with_generate": false,
|
| 158 |
+
"generation_max_length": null,
|
| 159 |
+
"generation_num_beams": null,
|
| 160 |
+
"generation_config": null,
|
| 161 |
+
"tuner_backend": "peft",
|
| 162 |
+
"vit_gradient_checkpointing": false,
|
| 163 |
+
"router_aux_loss_coef": 0.0,
|
| 164 |
+
"enable_dft_loss": false,
|
| 165 |
+
"enable_channel_loss": false,
|
| 166 |
+
"safe_serialization": true,
|
| 167 |
+
"max_shard_size": "5GB",
|
| 168 |
+
"check_model": false,
|
| 169 |
+
"acc_strategy": "token",
|
| 170 |
+
"train_dataloader_shuffle": true,
|
| 171 |
+
"group_by_length": false,
|
| 172 |
+
"max_epochs": null,
|
| 173 |
+
"aligner_lr": null,
|
| 174 |
+
"vit_lr": null,
|
| 175 |
+
"use_logits_to_keep": null,
|
| 176 |
+
"ds3_gather_for_generation": true,
|
| 177 |
+
"resume_only_model": false,
|
| 178 |
+
"optimizer": null,
|
| 179 |
+
"loss_type": null,
|
| 180 |
+
"mrl_dims": null,
|
| 181 |
+
"eval_metric": null,
|
| 182 |
+
"callbacks": [
|
| 183 |
+
"stop_at_step"
|
| 184 |
+
],
|
| 185 |
+
"early_stop_interval": null,
|
| 186 |
+
"eval_use_evalscope": false,
|
| 187 |
+
"eval_dataset": [],
|
| 188 |
+
"eval_dataset_args": null,
|
| 189 |
+
"eval_limit": null,
|
| 190 |
+
"eval_generation_config": null,
|
| 191 |
+
"extra_eval_args": null,
|
| 192 |
+
"tuner_type": "full",
|
| 193 |
+
"use_galore": false,
|
| 194 |
+
"galore_target_modules": null,
|
| 195 |
+
"galore_rank": 128,
|
| 196 |
+
"galore_update_proj_gap": 50,
|
| 197 |
+
"galore_scale": 1.0,
|
| 198 |
+
"galore_proj_type": "std",
|
| 199 |
+
"galore_optim_per_parameter": false,
|
| 200 |
+
"galore_with_embedding": false,
|
| 201 |
+
"galore_quantization": false,
|
| 202 |
+
"galore_proj_quant": false,
|
| 203 |
+
"galore_proj_bits": 4,
|
| 204 |
+
"galore_proj_group_size": 256,
|
| 205 |
+
"galore_cos_threshold": 0.4,
|
| 206 |
+
"galore_gamma_proj": 2,
|
| 207 |
+
"galore_queue_size": 5,
|
| 208 |
+
"lisa_activated_layers": 0,
|
| 209 |
+
"lisa_step_interval": 20,
|
| 210 |
+
"use_flash_ckpt": false,
|
| 211 |
+
"use_ray": false,
|
| 212 |
+
"ray_exp_name": null,
|
| 213 |
+
"device_groups": null,
|
| 214 |
+
"model": "/inspire/qb-ilm/project/semantic-visual-tokenizer/public/hym/weight/Qwen3Omni/models--Qwen--Qwen3-Omni-30B-A3B-Instruct/QwenOmni/snapshots/26291f793822fb6be9555850f06dfe95f2d7e695",
|
| 215 |
+
"model_type": "qwen3_omni_moe",
|
| 216 |
+
"model_revision": null,
|
| 217 |
+
"task_type": "causal_lm",
|
| 218 |
+
"torch_dtype": "bfloat16",
|
| 219 |
+
"attn_impl": "sdpa",
|
| 220 |
+
"experts_impl": "eager",
|
| 221 |
+
"new_special_tokens": [],
|
| 222 |
+
"num_labels": null,
|
| 223 |
+
"problem_type": null,
|
| 224 |
+
"rope_scaling": null,
|
| 225 |
+
"device_map": null,
|
| 226 |
+
"max_memory": {},
|
| 227 |
+
"max_model_len": null,
|
| 228 |
+
"local_repo_path": null,
|
| 229 |
+
"init_strategy": null,
|
| 230 |
+
"template": "qwen3_omni",
|
| 231 |
+
"system": null,
|
| 232 |
+
"max_length": 24576,
|
| 233 |
+
"truncation_strategy": "delete",
|
| 234 |
+
"max_pixels": null,
|
| 235 |
+
"agent_template": null,
|
| 236 |
+
"norm_bbox": null,
|
| 237 |
+
"use_chat_template": true,
|
| 238 |
+
"padding_side": "right",
|
| 239 |
+
"padding_free": false,
|
| 240 |
+
"loss_scale": "default",
|
| 241 |
+
"sequence_parallel_size": 1,
|
| 242 |
+
"is_binary_loss_scale": null,
|
| 243 |
+
"template_backend": "swift",
|
| 244 |
+
"response_prefix": null,
|
| 245 |
+
"enable_thinking": null,
|
| 246 |
+
"preserve_thinking": null,
|
| 247 |
+
"add_non_thinking_prefix": true,
|
| 248 |
+
"dataset": [
|
| 249 |
+
"/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run/train.abs_videos.jsonl"
|
| 250 |
+
],
|
| 251 |
+
"val_dataset": [],
|
| 252 |
+
"cached_dataset": [],
|
| 253 |
+
"cached_val_dataset": [],
|
| 254 |
+
"split_dataset_ratio": 0.0,
|
| 255 |
+
"dataset_num_proc": 4,
|
| 256 |
+
"load_from_cache_file": true,
|
| 257 |
+
"dataset_shuffle": true,
|
| 258 |
+
"val_dataset_shuffle": false,
|
| 259 |
+
"streaming": false,
|
| 260 |
+
"interleave_prob": null,
|
| 261 |
+
"stopping_strategy": "first_exhausted",
|
| 262 |
+
"shuffle_buffer_size": 1000,
|
| 263 |
+
"download_mode": "reuse_dataset_if_exists",
|
| 264 |
+
"columns": {},
|
| 265 |
+
"strict": false,
|
| 266 |
+
"disable_auto_column_mapping": false,
|
| 267 |
+
"model_name": null,
|
| 268 |
+
"model_author": null,
|
| 269 |
+
"custom_dataset_info": [],
|
| 270 |
+
"quant_method": null,
|
| 271 |
+
"quant_bits": null,
|
| 272 |
+
"hqq_axis": null,
|
| 273 |
+
"bnb_4bit_compute_dtype": "bfloat16",
|
| 274 |
+
"bnb_4bit_quant_type": "nf4",
|
| 275 |
+
"bnb_4bit_use_double_quant": true,
|
| 276 |
+
"bnb_4bit_quant_storage": null,
|
| 277 |
+
"max_new_tokens": 64,
|
| 278 |
+
"temperature": 0.0,
|
| 279 |
+
"top_k": null,
|
| 280 |
+
"top_p": null,
|
| 281 |
+
"repetition_penalty": null,
|
| 282 |
+
"num_beams": 1,
|
| 283 |
+
"stream": false,
|
| 284 |
+
"stop_words": [],
|
| 285 |
+
"logprobs": false,
|
| 286 |
+
"top_logprobs": null,
|
| 287 |
+
"structured_outputs_regex": null,
|
| 288 |
+
"adapters": [],
|
| 289 |
+
"external_plugins": [],
|
| 290 |
+
"custom_register_path": [],
|
| 291 |
+
"model_kwargs": {},
|
| 292 |
+
"enable_npu_model_patch": true,
|
| 293 |
+
"load_args": false,
|
| 294 |
+
"load_data_args": false,
|
| 295 |
+
"packing": false,
|
| 296 |
+
"packing_length": null,
|
| 297 |
+
"packing_num_proc": 1,
|
| 298 |
+
"lazy_tokenize": true,
|
| 299 |
+
"use_hf": false,
|
| 300 |
+
"ignore_args_error": false,
|
| 301 |
+
"use_swift_lora": false,
|
| 302 |
+
"freeze_parameters": [
|
| 303 |
+
"thinker.audio_tower",
|
| 304 |
+
"thinker.visual",
|
| 305 |
+
"thinker.audio_tower.proj1",
|
| 306 |
+
"thinker.audio_tower.proj2",
|
| 307 |
+
"thinker.visual.merger",
|
| 308 |
+
"thinker.visual.merger_list",
|
| 309 |
+
"talker",
|
| 310 |
+
"code2wav"
|
| 311 |
+
],
|
| 312 |
+
"freeze_parameters_regex": null,
|
| 313 |
+
"freeze_parameters_ratio": 0.0,
|
| 314 |
+
"trainable_parameters": [],
|
| 315 |
+
"trainable_parameters_regex": null,
|
| 316 |
+
"freeze_llm": false,
|
| 317 |
+
"freeze_vit": true,
|
| 318 |
+
"freeze_aligner": true,
|
| 319 |
+
"target_modules": [
|
| 320 |
+
"all-linear"
|
| 321 |
+
],
|
| 322 |
+
"target_regex": null,
|
| 323 |
+
"target_parameters": null,
|
| 324 |
+
"modules_to_save": [],
|
| 325 |
+
"lora_rank": 8,
|
| 326 |
+
"lora_alpha": 32,
|
| 327 |
+
"lora_dropout": 0.05,
|
| 328 |
+
"lora_bias": "none",
|
| 329 |
+
"lora_dtype": null,
|
| 330 |
+
"lorap_lr_ratio": null,
|
| 331 |
+
"use_rslora": false,
|
| 332 |
+
"use_dora": false,
|
| 333 |
+
"lora_ga_batch_size": 2,
|
| 334 |
+
"lora_ga_iters": 2,
|
| 335 |
+
"lora_ga_max_length": 1024,
|
| 336 |
+
"lora_ga_direction": "ArB2r",
|
| 337 |
+
"lora_ga_scale": "stable",
|
| 338 |
+
"lora_ga_stable_gamma": 16,
|
| 339 |
+
"init_weights": true,
|
| 340 |
+
"fourier_n_frequency": 2000,
|
| 341 |
+
"fourier_scaling": 300.0,
|
| 342 |
+
"boft_block_size": 4,
|
| 343 |
+
"boft_block_num": 0,
|
| 344 |
+
"boft_n_butterfly_factor": 1,
|
| 345 |
+
"boft_dropout": 0.0,
|
| 346 |
+
"vera_rank": 256,
|
| 347 |
+
"vera_projection_prng_key": 0,
|
| 348 |
+
"vera_dropout": 0.0,
|
| 349 |
+
"vera_d_initial": 0.1,
|
| 350 |
+
"adapter_act": "gelu",
|
| 351 |
+
"adapter_length": 128,
|
| 352 |
+
"adalora_target_r": 8,
|
| 353 |
+
"adalora_init_r": 12,
|
| 354 |
+
"adalora_tinit": 0,
|
| 355 |
+
"adalora_tfinal": 0,
|
| 356 |
+
"adalora_deltaT": 1,
|
| 357 |
+
"adalora_beta1": 0.85,
|
| 358 |
+
"adalora_beta2": 0.85,
|
| 359 |
+
"adalora_orth_reg_weight": 0.5,
|
| 360 |
+
"llamapro_num_new_blocks": 4,
|
| 361 |
+
"llamapro_num_groups": null,
|
| 362 |
+
"reft_layer_key": null,
|
| 363 |
+
"reft_layers": null,
|
| 364 |
+
"reft_rank": 4,
|
| 365 |
+
"reft_intervention_type": "LoreftIntervention",
|
| 366 |
+
"reft_args": null,
|
| 367 |
+
"swanlab_token": null,
|
| 368 |
+
"swanlab_project": "ms-swift",
|
| 369 |
+
"swanlab_workspace": null,
|
| 370 |
+
"swanlab_exp_name": null,
|
| 371 |
+
"swanlab_notification_method": null,
|
| 372 |
+
"swanlab_webhook_url": null,
|
| 373 |
+
"swanlab_secret": null,
|
| 374 |
+
"swanlab_sender_email": null,
|
| 375 |
+
"swanlab_receiver_email": null,
|
| 376 |
+
"swanlab_smtp_server": null,
|
| 377 |
+
"swanlab_smtp_port": null,
|
| 378 |
+
"swanlab_email_language": "zh",
|
| 379 |
+
"swanlab_mode": "cloud",
|
| 380 |
+
"add_version": false,
|
| 381 |
+
"create_checkpoint_symlink": false,
|
| 382 |
+
"zero_hpz_partition_size": null,
|
| 383 |
+
"deepspeed_autotp_size": null,
|
| 384 |
+
"swift_version": "4.4.0.dev0",
|
| 385 |
+
"ckpt_dir": null,
|
| 386 |
+
"rank": 0,
|
| 387 |
+
"global_world_size": 8,
|
| 388 |
+
"local_world_size": 8,
|
| 389 |
+
"model_suffix": "Qwen3-Omni-30B-A3B-Instruct/QwenOmni",
|
| 390 |
+
"model_info": "ModelInfo(model_type='qwen3_omni_moe', model_dir='/inspire/qb-ilm/project/semantic-visual-tokenizer/public/hym/weight/Qwen3Omni/models--Qwen--Qwen3-Omni-30B-A3B-Instruct/QwenOmni/snapshots/26291f793822fb6be9555850f06dfe95f2d7e695', torch_dtype=torch.bfloat16, max_model_len=65536, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=True, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)",
|
| 391 |
+
"model_meta": "ModelMeta(model_type='qwen3_omni_moe', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Instruct', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Thinking', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Captioner', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Captioner', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_omni', ignore_patterns=None, requires=None, tags=[])], loader=<class 'swift.model.models.qwen.Qwen3OmniLoader'>, template=None, model_arch=MultiModelKeys(arch_name='qwen3_omni', 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=['thinker.model', 'thinker.lm_head'], aligner=['thinker.audio_tower.proj1', 'thinker.audio_tower.proj2', 'thinker.visual.merger', 'thinker.visual.merger_list'], vision_tower=['thinker.audio_tower', 'thinker.visual'], generator=['talker', 'code2wav']), mcore_model_type=None, architectures=['Qwen3OmniMoeForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57.dev0', 'soundfile', 'decord', 'qwen_omni_utils>=0.0.9'], tags=['vision', 'video', 'audio'])",
|
| 392 |
+
"model_dir": "/inspire/qb-ilm/project/semantic-visual-tokenizer/public/hym/weight/Qwen3Omni/models--Qwen--Qwen3-Omni-30B-A3B-Instruct/QwenOmni/snapshots/26291f793822fb6be9555850f06dfe95f2d7e695",
|
| 393 |
+
"template_meta": "QwenTemplateMeta(template_type='qwen3_omni', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=<class 'swift.template.templates.qwen.Qwen3OmniTemplate'>, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='hermes', is_thinking=False, thinking_prefix='<think>\\n', non_thinking_prefix='', history_thinking_prefix='')",
|
| 394 |
+
"_val_dataset_exists": false,
|
| 395 |
+
"hub": "<class 'swift.hub.hub.MSHub'>",
|
| 396 |
+
"evaluation_strategy": "steps",
|
| 397 |
+
"training_args": "Seq2SeqTrainingArguments(output_dir='/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run', per_device_train_batch_size=4, num_train_epochs=3.0, max_steps=2000, learning_rate=5e-06, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, lr_scheduler_kwargs=None, warmup_steps=0.05, optim=<OptimizerNames.ADAMW_TORCH_FUSED: 'adamw_torch_fused'>, optim_args=None, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=4, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_steps=5, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['tensorboard'], run_name='/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=<IntervalStrategy.NO: 'no'>, eval_steps=100.0, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=100, save_on_each_node=False, save_total_limit=30, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, 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, dataloader_drop_last=False, dataloader_num_workers=4, dataloader_pin_memory=True, dataloader_persistent_workers=False, dataloader_prefetch_factor=2, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 3, 'offload_optimizer': {'device': 'cpu', 'pin_memory': True}, 'offload_param': {'device': 'cpu', 'pin_memory': True}, 'overlap_comm': False, 'contiguous_gradients': True, 'sub_group_size': 1000000000.0, 'reduce_bucket_size': 'auto', 'stage3_prefetch_bucket_size': 'auto', 'stage3_param_persistence_threshold': 'auto', 'stage3_max_live_parameters': 1000000000.0, 'stage3_max_reuse_distance': 1000000000.0, 'stage3_gather_16bit_weights_on_model_save': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=False, do_predict=False, resume_from_checkpoint='/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run/checkpoint-1100', warmup_ratio=0.05, logging_dir='/inspire/qb-ilm/project/semantic-visual-tokenizer/czxs253130297/ms-swift/output_sft_qwen3_omni_base_final_train_h200_8gpu_periodic_eval_full_cleandata/run/runs', local_rank=0, 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, safe_serialization=True, max_shard_size='5GB', check_model=False, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, 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, mrl_dims=None, eval_metric=None, callbacks=['stop_at_step'], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)"
|
| 398 |
+
}
|
VA-Judger/chat_template.jinja
ADDED
|
@@ -0,0 +1,122 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0].role == 'system' %}
|
| 4 |
+
{%- if messages[0].content is string %}
|
| 5 |
+
{{- messages[0].content }}
|
| 6 |
+
{%- else %}
|
| 7 |
+
{%- for content in messages[0].content %}
|
| 8 |
+
{%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
|
| 9 |
+
{{- "<|vision_start|><|image_pad|><|vision_end|>" }}
|
| 10 |
+
{%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %}
|
| 11 |
+
{{- "<|audio_start|><|audio_pad|><|audio_end|>" }}
|
| 12 |
+
{%- elif content.type == 'video' or 'video' in content %}
|
| 13 |
+
{{- "<|vision_start|><|video_pad|><|vision_end|>" }}
|
| 14 |
+
{%- elif content.type == 'text' %}
|
| 15 |
+
{{- content.text }}
|
| 16 |
+
{%- endif %}
|
| 17 |
+
{%- endfor %}
|
| 18 |
+
{%- endif %}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{{- '\n\n' }}
|
| 21 |
+
{{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 22 |
+
{%- for tool in tools %}
|
| 23 |
+
{{- "\n" }}
|
| 24 |
+
{{- tool | tojson }}
|
| 25 |
+
{%- endfor %}
|
| 26 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 27 |
+
{%- else %}
|
| 28 |
+
{%- if messages[0].role == 'system' %}
|
| 29 |
+
{%- if messages[0].content is string %}
|
| 30 |
+
{{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
|
| 31 |
+
{%- else %}
|
| 32 |
+
{%- for content in messages[0].content %}
|
| 33 |
+
{%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
|
| 34 |
+
{{- '<|im_start|>system\n' +"<|vision_start|><|image_pad|><|vision_end|>"+ '<|im_end|>\n' }}
|
| 35 |
+
{%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %}
|
| 36 |
+
{{- '<|im_start|>system\n' +"<|audio_start|><|audio_pad|><|audio_end|>"+ '<|im_end|>\n' }}
|
| 37 |
+
{%- elif content.type == 'video' or 'video' in content %}
|
| 38 |
+
{{- '<|im_start|>system\n' +"<|vision_start|><|video_pad|><|vision_end|>"+ '<|im_end|>\n' }}
|
| 39 |
+
{%- elif content.type == 'text' %}
|
| 40 |
+
{{- '<|im_start|>system\n' +content.text+ '<|im_end|>\n' }}
|
| 41 |
+
{%- endif %}
|
| 42 |
+
{%- endfor %}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{%- endif %}
|
| 45 |
+
{%- endif %}
|
| 46 |
+
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
|
| 47 |
+
{%- for message in messages[::-1] %}
|
| 48 |
+
{%- set index = (messages|length - 1) - loop.index0 %}
|
| 49 |
+
{%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
|
| 50 |
+
{%- set ns.multi_step_tool = false %}
|
| 51 |
+
{%- set ns.last_query_index = index %}
|
| 52 |
+
{%- endif %}
|
| 53 |
+
{%- endfor %}
|
| 54 |
+
{%- for message in messages %}
|
| 55 |
+
{%- if message.content is string %}
|
| 56 |
+
{%- set content = message.content %}
|
| 57 |
+
{%- else %}
|
| 58 |
+
{%- set content = namespace(text="") %}
|
| 59 |
+
{%- for mcontent in message.content %}
|
| 60 |
+
{%- if mcontent.type == 'image' or 'image' in mcontent or 'image_url' in mcontent %}
|
| 61 |
+
{%- set content.text = content.text~"<|vision_start|><|image_pad|><|vision_end|>" %}
|
| 62 |
+
{%- elif mcontent.type == 'audio' or 'audio' in mcontent or 'audio_url' in mcontent %}
|
| 63 |
+
{%- set content.text = content.text~"<|audio_start|><|audio_pad|><|audio_end|>" %}
|
| 64 |
+
{%- elif mcontent.type == 'video' or 'video' in mcontent %}
|
| 65 |
+
{%- set content.text = content.text~"<|vision_start|><|video_pad|><|vision_end|>" %}
|
| 66 |
+
{%- elif mcontent.type == 'text' %}
|
| 67 |
+
{%- set content.text = content.text~mcontent.text %}
|
| 68 |
+
{%- endif %}
|
| 69 |
+
{%- endfor %}
|
| 70 |
+
{%- set content = content.text %}
|
| 71 |
+
{%- endif %}
|
| 72 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
|
| 73 |
+
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
|
| 74 |
+
{%- elif message.role == "assistant" %}
|
| 75 |
+
{%- set reasoning_content = "" %}
|
| 76 |
+
{%- if message.reasoning_content is string %}
|
| 77 |
+
{%- set reasoning_content = message.reasoning_content %}
|
| 78 |
+
{%- else %}
|
| 79 |
+
{%- if '</think>' in content %}
|
| 80 |
+
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
|
| 81 |
+
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
|
| 82 |
+
{%- endif %}
|
| 83 |
+
{%- endif %}
|
| 84 |
+
{%- if loop.index0 > ns.last_query_index %}
|
| 85 |
+
{%- if loop.last or (not loop.last and reasoning_content) %}
|
| 86 |
+
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip("\n") + '\n</think>\n\n' + content.lstrip('\n') }}
|
| 87 |
+
{%- else %}
|
| 88 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 89 |
+
{%- endif %}
|
| 90 |
+
{%- else %}
|
| 91 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 92 |
+
{%- endif %}
|
| 93 |
+
{%- if message.tool_calls %}
|
| 94 |
+
{%- for tool_call in message.tool_calls %}
|
| 95 |
+
{%- if (loop.first and content) or (not loop.first) %}{{- '\n' }}{%- endif %}
|
| 96 |
+
{%- if tool_call.function %}
|
| 97 |
+
{%- set tool_call = tool_call.function %}
|
| 98 |
+
{%- endif %}
|
| 99 |
+
{{- '<tool_call>\n{"name": "' }}
|
| 100 |
+
{{- tool_call.name }}
|
| 101 |
+
{{- '", "arguments": ' }}
|
| 102 |
+
{%- if tool_call.arguments is string %}
|
| 103 |
+
{{- tool_call.arguments }}
|
| 104 |
+
{%- else %}
|
| 105 |
+
{{- tool_call.arguments | tojson }}
|
| 106 |
+
{%- endif %}
|
| 107 |
+
{{- '}\n</tool_call>' }}
|
| 108 |
+
{%- endfor %}
|
| 109 |
+
{%- endif %}
|
| 110 |
+
{{- '<|im_end|>\n' }}
|
| 111 |
+
{%- elif message.role == "tool" %}
|
| 112 |
+
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}{{- '<|im_start|>user' }}{%- endif %}
|
| 113 |
+
{{- '\n<tool_response>\n' }}
|
| 114 |
+
{{- content }}
|
| 115 |
+
{{- '\n</tool_response>' }}
|
| 116 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}{{- '<|im_end|>\n' }}{%- endif %}
|
| 117 |
+
{%- endif %}
|
| 118 |
+
{%- endfor %}
|
| 119 |
+
{%- if add_generation_prompt %}
|
| 120 |
+
{{- '<|im_start|>assistant\n' }}
|
| 121 |
+
{%- if enable_thinking is defined and enable_thinking is false %}{{- '<think>\n\n</think>\n\n' }}{%- endif %}
|
| 122 |
+
{%- endif %}
|
VA-Judger/chat_template.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0].role == 'system' %}\n {%- if messages[0].content is string %}\n {{- messages[0].content }}\n {%- else %}\n {%- for content in messages[0].content %}\n {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}\n {{- \"<|vision_start|><|image_pad|><|vision_end|>\" }}\n {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %}\n {{- \"<|audio_start|><|audio_pad|><|audio_end|>\" }}\n {%- elif content.type == 'video' or 'video' in content %}\n {{- \"<|vision_start|><|video_pad|><|vision_end|>\" }}\n {%- elif content.type == 'text' %}\n {{- content.text }}\n {%- endif %}\n {%- endfor %}\n {%- endif %}\n {%- endif %}\n {{- '\\n\\n' }}\n {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {%- if messages[0].content is string %}\n {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n {%- else %}\n {%- for content in messages[0].content %}\n {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}\n {{- '<|im_start|>system\\n' +\"<|vision_start|><|image_pad|><|vision_end|>\"+ '<|im_end|>\\n' }}\n {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %}\n {{- '<|im_start|>system\\n' +\"<|audio_start|><|audio_pad|><|audio_end|>\"+ '<|im_end|>\\n' }}\n {%- elif content.type == 'video' or 'video' in content %}\n {{- '<|im_start|>system\\n' +\"<|vision_start|><|video_pad|><|vision_end|>\"+ '<|im_end|>\\n' }}\n {%- elif content.type == 'text' %}\n {{- '<|im_start|>system\\n' +content.text+ '<|im_end|>\\n' }}\n {%- endif %}\n {%- endfor %}\n {%- endif %}\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\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n {%- if message.content is string %}\n {%- set content = message.content %}\n {%- else %}\n {%- set content = namespace(text=\"\") %}\n {%- for mcontent in message.content %}\n {%- if mcontent.type == 'image' or 'image' in mcontent or 'image_url' in mcontent %}\n {%- set content.text = content.text~\"<|vision_start|><|image_pad|><|vision_end|>\" %}\n {%- elif mcontent.type == 'audio' or 'audio' in mcontent or 'audio_url' in mcontent %}\n {%- set content.text = content.text~\"<|audio_start|><|audio_pad|><|audio_end|>\" %}\n {%- elif mcontent.type == 'video' or 'video' in mcontent %}\n {%- set content.text = content.text~\"<|vision_start|><|video_pad|><|vision_end|>\" %}\n {%- elif mcontent.type == 'text' %}\n {%- set content.text = content.text~mcontent.text %}\n {%- endif %}\n {%- endfor %}\n {%- set content = content.text %}\n {%- endif %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\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 {%- if loop.index0 > ns.last_query_index %}\n {%- if loop.last or (not loop.last and reasoning_content) %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip(\"\\n\") + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls %}\n {%- for tool_call in message.tool_calls %}\n {%- if (loop.first and content) or (not loop.first) %}{{- '\\n' }}{%- endif %}\n {%- if tool_call.function %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {%- if tool_call.arguments is string %}\n {{- tool_call.arguments }}\n {%- else %}\n {{- tool_call.arguments | tojson }}\n {%- endif %}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}{{- '<|im_start|>user' }}{%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}{{- '<|im_end|>\\n' }}{%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is false %}{{- '<think>\\n\\n</think>\\n\\n' }}{%- endif %}\n{%- endif %}"}
|
VA-Judger/config.json
ADDED
|
@@ -0,0 +1,430 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3OmniMoeForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"assistant_token_id": 77091,
|
| 6 |
+
"code2wav_config": {
|
| 7 |
+
"attention_bias": false,
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"codebook_dim": 512,
|
| 10 |
+
"codebook_size": 2048,
|
| 11 |
+
"decoder_dim": 1536,
|
| 12 |
+
"dtype": "bfloat16",
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 1024,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 3072,
|
| 17 |
+
"layer_scale_initial_scale": 0.01,
|
| 18 |
+
"max_position_embeddings": 8000,
|
| 19 |
+
"model_type": "",
|
| 20 |
+
"num_attention_heads": 16,
|
| 21 |
+
"num_hidden_layers": 8,
|
| 22 |
+
"num_key_value_heads": 16,
|
| 23 |
+
"num_quantizers": 16,
|
| 24 |
+
"num_semantic_quantizers": 1,
|
| 25 |
+
"rms_norm_eps": 1e-05,
|
| 26 |
+
"rope_parameters": {
|
| 27 |
+
"rope_theta": 10000,
|
| 28 |
+
"rope_type": "default"
|
| 29 |
+
},
|
| 30 |
+
"semantic_codebook_size": 4096,
|
| 31 |
+
"sliding_window": 72,
|
| 32 |
+
"upsample_rates": [
|
| 33 |
+
8,
|
| 34 |
+
5,
|
| 35 |
+
4,
|
| 36 |
+
3
|
| 37 |
+
],
|
| 38 |
+
"upsampling_ratios": [
|
| 39 |
+
2,
|
| 40 |
+
2
|
| 41 |
+
],
|
| 42 |
+
"vector_quantization_hidden_dimension": 512
|
| 43 |
+
},
|
| 44 |
+
"dtype": "bfloat16",
|
| 45 |
+
"enable_audio_output": false,
|
| 46 |
+
"eos_token_id": 151645,
|
| 47 |
+
"hidden_size": 2048,
|
| 48 |
+
"im_end_token_id": 151645,
|
| 49 |
+
"im_start_token_id": 151644,
|
| 50 |
+
"initializer_range": 0.02,
|
| 51 |
+
"keys_to_ignore_at_inference": [
|
| 52 |
+
"past_key_values",
|
| 53 |
+
"hidden_states",
|
| 54 |
+
"attention_mask"
|
| 55 |
+
],
|
| 56 |
+
"model_type": "qwen3_omni_moe",
|
| 57 |
+
"pad_token_id": 151643,
|
| 58 |
+
"system_token_id": 8948,
|
| 59 |
+
"talker_config": {
|
| 60 |
+
"accept_hidden_layer": 24,
|
| 61 |
+
"audio_end_token_id": 151670,
|
| 62 |
+
"audio_start_token_id": 151669,
|
| 63 |
+
"audio_token_id": 151675,
|
| 64 |
+
"code_predictor_config": {
|
| 65 |
+
"_name_or_path": "",
|
| 66 |
+
"add_cross_attention": false,
|
| 67 |
+
"architectures": null,
|
| 68 |
+
"attention_bias": false,
|
| 69 |
+
"attention_dropout": 0,
|
| 70 |
+
"bos_token_id": null,
|
| 71 |
+
"chunk_size_feed_forward": 0,
|
| 72 |
+
"cross_attention_hidden_size": null,
|
| 73 |
+
"decoder_start_token_id": null,
|
| 74 |
+
"dtype": null,
|
| 75 |
+
"eos_token_id": null,
|
| 76 |
+
"finetuning_task": null,
|
| 77 |
+
"head_dim": 128,
|
| 78 |
+
"hidden_act": "silu",
|
| 79 |
+
"hidden_size": 1024,
|
| 80 |
+
"id2label": {
|
| 81 |
+
"0": "LABEL_0",
|
| 82 |
+
"1": "LABEL_1"
|
| 83 |
+
},
|
| 84 |
+
"initializer_range": 0.02,
|
| 85 |
+
"intermediate_size": 3072,
|
| 86 |
+
"is_decoder": false,
|
| 87 |
+
"is_encoder_decoder": false,
|
| 88 |
+
"label2id": {
|
| 89 |
+
"LABEL_0": 0,
|
| 90 |
+
"LABEL_1": 1
|
| 91 |
+
},
|
| 92 |
+
"layer_types": [
|
| 93 |
+
"full_attention",
|
| 94 |
+
"full_attention",
|
| 95 |
+
"full_attention",
|
| 96 |
+
"full_attention",
|
| 97 |
+
"full_attention"
|
| 98 |
+
],
|
| 99 |
+
"max_position_embeddings": 32768,
|
| 100 |
+
"max_window_layers": 28,
|
| 101 |
+
"model_type": "qwen3_omni_moe_talker_code_predictor",
|
| 102 |
+
"num_attention_heads": 16,
|
| 103 |
+
"num_code_groups": 16,
|
| 104 |
+
"num_hidden_layers": 5,
|
| 105 |
+
"num_key_value_heads": 8,
|
| 106 |
+
"output_attentions": false,
|
| 107 |
+
"output_hidden_states": false,
|
| 108 |
+
"pad_token_id": null,
|
| 109 |
+
"prefix": null,
|
| 110 |
+
"problem_type": null,
|
| 111 |
+
"pruned_heads": {},
|
| 112 |
+
"return_dict": true,
|
| 113 |
+
"rms_norm_eps": 1e-06,
|
| 114 |
+
"rope_parameters": {
|
| 115 |
+
"rope_theta": 1000000,
|
| 116 |
+
"rope_type": "default"
|
| 117 |
+
},
|
| 118 |
+
"sep_token_id": null,
|
| 119 |
+
"sliding_window": null,
|
| 120 |
+
"task_specific_params": null,
|
| 121 |
+
"tf_legacy_loss": false,
|
| 122 |
+
"tie_encoder_decoder": false,
|
| 123 |
+
"tie_word_embeddings": false,
|
| 124 |
+
"tokenizer_class": null,
|
| 125 |
+
"torchscript": false,
|
| 126 |
+
"use_bfloat16": false,
|
| 127 |
+
"use_cache": true,
|
| 128 |
+
"use_sliding_window": false,
|
| 129 |
+
"vocab_size": 2048
|
| 130 |
+
},
|
| 131 |
+
"codec_bos_id": 2149,
|
| 132 |
+
"codec_eos_token_id": 2150,
|
| 133 |
+
"codec_nothink_id": 2155,
|
| 134 |
+
"codec_pad_id": 2148,
|
| 135 |
+
"codec_think_bos_id": 2156,
|
| 136 |
+
"codec_think_eos_id": 2157,
|
| 137 |
+
"dtype": "bfloat16",
|
| 138 |
+
"image_token_id": 151655,
|
| 139 |
+
"initializer_range": 0.02,
|
| 140 |
+
"model_type": "",
|
| 141 |
+
"num_code_groups": 16,
|
| 142 |
+
"output_router_logits": false,
|
| 143 |
+
"pad_token_id": null,
|
| 144 |
+
"position_id_per_seconds": 13,
|
| 145 |
+
"seconds_per_chunk": 2,
|
| 146 |
+
"spatial_merge_size": 2,
|
| 147 |
+
"speaker_id": {
|
| 148 |
+
"aiden": 2303,
|
| 149 |
+
"chelsie": 2301,
|
| 150 |
+
"ethan": 2302
|
| 151 |
+
},
|
| 152 |
+
"text_config": {
|
| 153 |
+
"_name_or_path": "",
|
| 154 |
+
"architectures": null,
|
| 155 |
+
"attention_bias": false,
|
| 156 |
+
"attention_dropout": 0,
|
| 157 |
+
"bos_token_id": null,
|
| 158 |
+
"chunk_size_feed_forward": 0,
|
| 159 |
+
"decoder_sparse_step": 1,
|
| 160 |
+
"dtype": null,
|
| 161 |
+
"eos_token_id": null,
|
| 162 |
+
"head_dim": 128,
|
| 163 |
+
"hidden_act": "silu",
|
| 164 |
+
"hidden_size": 1024,
|
| 165 |
+
"id2label": {
|
| 166 |
+
"0": "LABEL_0",
|
| 167 |
+
"1": "LABEL_1"
|
| 168 |
+
},
|
| 169 |
+
"initializer_range": 0.02,
|
| 170 |
+
"intermediate_size": 2048,
|
| 171 |
+
"is_encoder_decoder": false,
|
| 172 |
+
"label2id": {
|
| 173 |
+
"LABEL_0": 0,
|
| 174 |
+
"LABEL_1": 1
|
| 175 |
+
},
|
| 176 |
+
"max_position_embeddings": 65536,
|
| 177 |
+
"mlp_only_layers": [],
|
| 178 |
+
"model_type": "qwen3_omni_moe_talker_text",
|
| 179 |
+
"moe_intermediate_size": 384,
|
| 180 |
+
"norm_topk_prob": true,
|
| 181 |
+
"num_attention_heads": 16,
|
| 182 |
+
"num_experts_per_tok": 6,
|
| 183 |
+
"num_hidden_layers": 20,
|
| 184 |
+
"num_key_value_heads": 2,
|
| 185 |
+
"num_local_experts": 128,
|
| 186 |
+
"output_attentions": false,
|
| 187 |
+
"output_hidden_states": false,
|
| 188 |
+
"output_router_logits": false,
|
| 189 |
+
"pad_token_id": null,
|
| 190 |
+
"problem_type": null,
|
| 191 |
+
"return_dict": true,
|
| 192 |
+
"rms_norm_eps": 1e-06,
|
| 193 |
+
"rope_parameters": {
|
| 194 |
+
"interleaved": true,
|
| 195 |
+
"mrope_section": [
|
| 196 |
+
24,
|
| 197 |
+
20,
|
| 198 |
+
20
|
| 199 |
+
],
|
| 200 |
+
"rope_theta": 1000000,
|
| 201 |
+
"rope_type": "default",
|
| 202 |
+
"type": "default"
|
| 203 |
+
},
|
| 204 |
+
"router_aux_loss_coef": 0.001,
|
| 205 |
+
"shared_expert_intermediate_size": 768,
|
| 206 |
+
"sliding_window": null,
|
| 207 |
+
"tie_word_embeddings": false,
|
| 208 |
+
"use_cache": true,
|
| 209 |
+
"use_sliding_window": false,
|
| 210 |
+
"vocab_size": 3072
|
| 211 |
+
},
|
| 212 |
+
"thinker_hidden_size": 2048,
|
| 213 |
+
"tie_word_embeddings": false,
|
| 214 |
+
"video_token_id": 151656,
|
| 215 |
+
"vision_start_token_id": 151652
|
| 216 |
+
},
|
| 217 |
+
"thinker_config": {
|
| 218 |
+
"audio_config": {
|
| 219 |
+
"_name_or_path": "",
|
| 220 |
+
"activation_dropout": 0,
|
| 221 |
+
"activation_function": "gelu",
|
| 222 |
+
"add_cross_attention": false,
|
| 223 |
+
"architectures": null,
|
| 224 |
+
"attention_dropout": 0,
|
| 225 |
+
"bos_token_id": null,
|
| 226 |
+
"chunk_size_feed_forward": 0,
|
| 227 |
+
"conv_chunksize": 500,
|
| 228 |
+
"cross_attention_hidden_size": null,
|
| 229 |
+
"d_model": 1280,
|
| 230 |
+
"decoder_start_token_id": null,
|
| 231 |
+
"downsample_hidden_size": 480,
|
| 232 |
+
"dropout": 0,
|
| 233 |
+
"dtype": "bfloat16",
|
| 234 |
+
"encoder_attention_heads": 20,
|
| 235 |
+
"encoder_ffn_dim": 5120,
|
| 236 |
+
"encoder_layers": 32,
|
| 237 |
+
"eos_token_id": null,
|
| 238 |
+
"finetuning_task": null,
|
| 239 |
+
"id2label": {
|
| 240 |
+
"0": "LABEL_0",
|
| 241 |
+
"1": "LABEL_1"
|
| 242 |
+
},
|
| 243 |
+
"initializer_range": 0.02,
|
| 244 |
+
"is_decoder": false,
|
| 245 |
+
"is_encoder_decoder": false,
|
| 246 |
+
"label2id": {
|
| 247 |
+
"LABEL_0": 0,
|
| 248 |
+
"LABEL_1": 1
|
| 249 |
+
},
|
| 250 |
+
"max_source_positions": 1500,
|
| 251 |
+
"model_type": "qwen3_omni_moe_audio_encoder",
|
| 252 |
+
"n_window": 50,
|
| 253 |
+
"n_window_infer": 800,
|
| 254 |
+
"num_mel_bins": 128,
|
| 255 |
+
"output_attentions": false,
|
| 256 |
+
"output_dim": 2048,
|
| 257 |
+
"output_hidden_states": false,
|
| 258 |
+
"pad_token_id": null,
|
| 259 |
+
"prefix": null,
|
| 260 |
+
"problem_type": null,
|
| 261 |
+
"pruned_heads": {},
|
| 262 |
+
"return_dict": true,
|
| 263 |
+
"scale_embedding": false,
|
| 264 |
+
"sep_token_id": null,
|
| 265 |
+
"task_specific_params": null,
|
| 266 |
+
"tf_legacy_loss": false,
|
| 267 |
+
"tie_encoder_decoder": false,
|
| 268 |
+
"tie_word_embeddings": true,
|
| 269 |
+
"tokenizer_class": null,
|
| 270 |
+
"torchscript": false,
|
| 271 |
+
"use_bfloat16": false
|
| 272 |
+
},
|
| 273 |
+
"audio_end_token_id": 151670,
|
| 274 |
+
"audio_start_token_id": 151669,
|
| 275 |
+
"audio_token_id": 151675,
|
| 276 |
+
"dtype": "bfloat16",
|
| 277 |
+
"image_token_id": 151655,
|
| 278 |
+
"initializer_range": 0.02,
|
| 279 |
+
"model_type": "qwen3_omni_moe_thinker",
|
| 280 |
+
"position_id_per_seconds": 13,
|
| 281 |
+
"seconds_per_chunk": 2,
|
| 282 |
+
"text_config": {
|
| 283 |
+
"_name_or_path": "",
|
| 284 |
+
"add_cross_attention": false,
|
| 285 |
+
"architectures": null,
|
| 286 |
+
"attention_bias": false,
|
| 287 |
+
"attention_dropout": 0.0,
|
| 288 |
+
"bos_token_id": null,
|
| 289 |
+
"chunk_size_feed_forward": 0,
|
| 290 |
+
"cross_attention_hidden_size": null,
|
| 291 |
+
"decoder_sparse_step": 1,
|
| 292 |
+
"decoder_start_token_id": null,
|
| 293 |
+
"dtype": "bfloat16",
|
| 294 |
+
"eos_token_id": null,
|
| 295 |
+
"finetuning_task": null,
|
| 296 |
+
"head_dim": 128,
|
| 297 |
+
"hidden_act": "silu",
|
| 298 |
+
"hidden_size": 2048,
|
| 299 |
+
"id2label": {
|
| 300 |
+
"0": "LABEL_0",
|
| 301 |
+
"1": "LABEL_1"
|
| 302 |
+
},
|
| 303 |
+
"initializer_range": 0.02,
|
| 304 |
+
"intermediate_size": 768,
|
| 305 |
+
"is_decoder": false,
|
| 306 |
+
"is_encoder_decoder": false,
|
| 307 |
+
"label2id": {
|
| 308 |
+
"LABEL_0": 0,
|
| 309 |
+
"LABEL_1": 1
|
| 310 |
+
},
|
| 311 |
+
"max_position_embeddings": 65536,
|
| 312 |
+
"mlp_only_layers": [],
|
| 313 |
+
"model_type": "qwen3_omni_moe_text",
|
| 314 |
+
"moe_intermediate_size": 768,
|
| 315 |
+
"norm_topk_prob": true,
|
| 316 |
+
"num_attention_heads": 32,
|
| 317 |
+
"num_experts": 128,
|
| 318 |
+
"num_experts_per_tok": 8,
|
| 319 |
+
"num_hidden_layers": 48,
|
| 320 |
+
"num_key_value_heads": 4,
|
| 321 |
+
"output_attentions": false,
|
| 322 |
+
"output_hidden_states": false,
|
| 323 |
+
"output_router_logits": false,
|
| 324 |
+
"pad_token_id": 151643,
|
| 325 |
+
"prefix": null,
|
| 326 |
+
"problem_type": null,
|
| 327 |
+
"pruned_heads": {},
|
| 328 |
+
"return_dict": true,
|
| 329 |
+
"rms_norm_eps": 1e-06,
|
| 330 |
+
"rope_parameters": {
|
| 331 |
+
"interleaved": true,
|
| 332 |
+
"mrope_interleaved": true,
|
| 333 |
+
"mrope_section": [
|
| 334 |
+
24,
|
| 335 |
+
20,
|
| 336 |
+
20
|
| 337 |
+
],
|
| 338 |
+
"rope_theta": 1000000,
|
| 339 |
+
"rope_type": "default",
|
| 340 |
+
"type": "default"
|
| 341 |
+
},
|
| 342 |
+
"router_aux_loss_coef": 0.0,
|
| 343 |
+
"sep_token_id": null,
|
| 344 |
+
"shared_expert_intermediate_size": 0,
|
| 345 |
+
"sliding_window": null,
|
| 346 |
+
"task_specific_params": null,
|
| 347 |
+
"tf_legacy_loss": false,
|
| 348 |
+
"tie_encoder_decoder": false,
|
| 349 |
+
"tie_word_embeddings": false,
|
| 350 |
+
"tokenizer_class": null,
|
| 351 |
+
"torchscript": false,
|
| 352 |
+
"use_bfloat16": false,
|
| 353 |
+
"use_cache": false,
|
| 354 |
+
"use_qk_norm": true,
|
| 355 |
+
"use_sliding_window": false,
|
| 356 |
+
"vocab_size": 152064
|
| 357 |
+
},
|
| 358 |
+
"tie_word_embeddings": false,
|
| 359 |
+
"user_token_id": 872,
|
| 360 |
+
"video_token_id": 151656,
|
| 361 |
+
"vision_config": {
|
| 362 |
+
"_name_or_path": "",
|
| 363 |
+
"add_cross_attention": false,
|
| 364 |
+
"apply_vit_abs_pos_embed": true,
|
| 365 |
+
"architectures": null,
|
| 366 |
+
"bos_token_id": null,
|
| 367 |
+
"chunk_size_feed_forward": 0,
|
| 368 |
+
"cross_attention_hidden_size": null,
|
| 369 |
+
"decoder_start_token_id": null,
|
| 370 |
+
"deepstack_visual_indexes": [
|
| 371 |
+
8,
|
| 372 |
+
16,
|
| 373 |
+
24
|
| 374 |
+
],
|
| 375 |
+
"depth": 27,
|
| 376 |
+
"dtype": "bfloat16",
|
| 377 |
+
"eos_token_id": null,
|
| 378 |
+
"finetuning_task": null,
|
| 379 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 380 |
+
"hidden_size": 1152,
|
| 381 |
+
"id2label": {
|
| 382 |
+
"0": "LABEL_0",
|
| 383 |
+
"1": "LABEL_1"
|
| 384 |
+
},
|
| 385 |
+
"image_size": 768,
|
| 386 |
+
"in_channels": 3,
|
| 387 |
+
"in_chans": 3,
|
| 388 |
+
"initializer_range": 0.02,
|
| 389 |
+
"intermediate_size": 4304,
|
| 390 |
+
"is_decoder": false,
|
| 391 |
+
"is_encoder_decoder": false,
|
| 392 |
+
"label2id": {
|
| 393 |
+
"LABEL_0": 0,
|
| 394 |
+
"LABEL_1": 1
|
| 395 |
+
},
|
| 396 |
+
"model_type": "qwen3_omni_moe_vision_encoder",
|
| 397 |
+
"num_heads": 16,
|
| 398 |
+
"num_position_embeddings": 2304,
|
| 399 |
+
"out_hidden_size": 2048,
|
| 400 |
+
"output_attentions": false,
|
| 401 |
+
"output_hidden_states": false,
|
| 402 |
+
"pad_token_id": null,
|
| 403 |
+
"patch_size": 16,
|
| 404 |
+
"prefix": null,
|
| 405 |
+
"problem_type": null,
|
| 406 |
+
"pruned_heads": {},
|
| 407 |
+
"return_dict": true,
|
| 408 |
+
"sep_token_id": null,
|
| 409 |
+
"spatial_merge_size": 2,
|
| 410 |
+
"spatial_patch_size": 16,
|
| 411 |
+
"task_specific_params": null,
|
| 412 |
+
"temporal_patch_size": 2,
|
| 413 |
+
"tf_legacy_loss": false,
|
| 414 |
+
"tie_encoder_decoder": false,
|
| 415 |
+
"tie_word_embeddings": true,
|
| 416 |
+
"tokenizer_class": null,
|
| 417 |
+
"tokens_per_second": 2,
|
| 418 |
+
"torchscript": false,
|
| 419 |
+
"use_bfloat16": false
|
| 420 |
+
},
|
| 421 |
+
"vision_end_token_id": 151653,
|
| 422 |
+
"vision_start_token_id": 151652
|
| 423 |
+
},
|
| 424 |
+
"transformers_version": "5.10.2",
|
| 425 |
+
"tts_bos_token_id": 151672,
|
| 426 |
+
"tts_eos_token_id": 151673,
|
| 427 |
+
"tts_pad_token_id": 151671,
|
| 428 |
+
"use_cache": false,
|
| 429 |
+
"user_token_id": 872
|
| 430 |
+
}
|
VA-Judger/generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token_id": [
|
| 3 |
+
151645,
|
| 4 |
+
151643
|
| 5 |
+
],
|
| 6 |
+
"talker_max_new_tokens": 4096,
|
| 7 |
+
"talker_repetition_penalty": 1.05,
|
| 8 |
+
"talker_temperature": 0.9,
|
| 9 |
+
"talker_top_k": 50,
|
| 10 |
+
"talker_top_p": 1.0,
|
| 11 |
+
"transformers_version": "5.10.2"
|
| 12 |
+
}
|
VA-Judger/latest
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
global_step1200
|
VA-Judger/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
VA-Judger/model-00001-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0c7b3507ba79d525c54a93f280e42ec579660e256514d469781d18d55407912
|
| 3 |
+
size 4998560832
|
VA-Judger/model-00002-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae6d888310317f5e9b02eea3b616fee38a5be4d1c43b946dd710fe51d61b12b1
|
| 3 |
+
size 4985170824
|
VA-Judger/model-00003-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:567ac1a1298a8235da4193bbbde029bde4289c41226d5506f2c06c827e23b8c9
|
| 3 |
+
size 4985170824
|
VA-Judger/model-00004-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0d214e306023cd18d6c166d3ee9ae1f5546be1384bdec8635bcf8350218568f
|
| 3 |
+
size 4985172352
|
VA-Judger/model-00005-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86e30084bfce41ec8f5f78433d80d69b5d551faf62cc2174ee2db40fb631f522
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00006-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb53e6fb7a4335af01c8c70b7770b786b2758af795bb6dcafe16c01e0a848e2c
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00007-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ee5caec0f1db3e2a5a535436259c8bb8e6eb90fac895e1ca7d5baa7b1d51c49
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00008-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35adcba037f4912ca0d8744923b8d0757c399e4dbf0a7ee36a70a983bf8e9c6c
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00009-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36ec6f70b35483d9b998d7a6686c6f636e505fba6cca8dfbe5dba4dcfa680657
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00010-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72387a0a96bbe08eb4bf92683e908a3240f330c92362c849406a9fb26f712aec
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00011-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74142ca64876cb1e8b378004d568b65e34fc7d9fa8a1ab254d0b33104fe62999
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00012-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:903806acf6c1c505e3efce0bcc41bcc9c4fbbd4f27d03aa6aa2aa2517d37c564
|
| 3 |
+
size 4985172392
|
VA-Judger/model-00013-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37a043f69f85414600f8d50775c72bb8b80e33db5eae41aca85277c87cd014a5
|
| 3 |
+
size 3605545704
|
VA-Judger/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
VA-Judger/preprocessor_config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dither": 0.0,
|
| 3 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 4 |
+
"feature_size": 128,
|
| 5 |
+
"hop_length": 160,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "Qwen2VLImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"max_pixels": 12845056,
|
| 18 |
+
"merge_size": 2,
|
| 19 |
+
"min_pixels": 3136,
|
| 20 |
+
"n_fft": 400,
|
| 21 |
+
"n_samples": 4800000,
|
| 22 |
+
"nb_max_frames": 30000,
|
| 23 |
+
"padding_side": "right",
|
| 24 |
+
"padding_value": 0.0,
|
| 25 |
+
"patch_size": 16,
|
| 26 |
+
"processor_class": "Qwen3OmniMoeProcessor",
|
| 27 |
+
"return_attention_mask": true,
|
| 28 |
+
"sampling_rate": 16000,
|
| 29 |
+
"temporal_patch_size": 2
|
| 30 |
+
}
|
VA-Judger/processor_config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"feature_extractor": {
|
| 3 |
+
"chunk_length": 30,
|
| 4 |
+
"dither": 0.0,
|
| 5 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 6 |
+
"feature_size": 128,
|
| 7 |
+
"hop_length": 160,
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.5,
|
| 10 |
+
0.5,
|
| 11 |
+
0.5
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "Qwen2VLImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.5,
|
| 16 |
+
0.5,
|
| 17 |
+
0.5
|
| 18 |
+
],
|
| 19 |
+
"max_pixels": 12845056,
|
| 20 |
+
"merge_size": 2,
|
| 21 |
+
"min_pixels": 3136,
|
| 22 |
+
"n_fft": 400,
|
| 23 |
+
"n_samples": 480000,
|
| 24 |
+
"nb_max_frames": 3000,
|
| 25 |
+
"padding_side": "right",
|
| 26 |
+
"padding_value": 0.0,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"return_attention_mask": true,
|
| 29 |
+
"sampling_rate": 16000,
|
| 30 |
+
"temporal_patch_size": 2
|
| 31 |
+
},
|
| 32 |
+
"image_processor": {
|
| 33 |
+
"dither": 0.0,
|
| 34 |
+
"do_convert_rgb": true,
|
| 35 |
+
"do_normalize": true,
|
| 36 |
+
"do_rescale": true,
|
| 37 |
+
"do_resize": true,
|
| 38 |
+
"feature_size": 128,
|
| 39 |
+
"hop_length": 160,
|
| 40 |
+
"image_mean": [
|
| 41 |
+
0.5,
|
| 42 |
+
0.5,
|
| 43 |
+
0.5
|
| 44 |
+
],
|
| 45 |
+
"image_processor_type": "Qwen2VLImageProcessor",
|
| 46 |
+
"image_std": [
|
| 47 |
+
0.5,
|
| 48 |
+
0.5,
|
| 49 |
+
0.5
|
| 50 |
+
],
|
| 51 |
+
"merge_size": 2,
|
| 52 |
+
"n_fft": 400,
|
| 53 |
+
"n_samples": 4800000,
|
| 54 |
+
"nb_max_frames": 30000,
|
| 55 |
+
"padding_side": "right",
|
| 56 |
+
"padding_value": 0.0,
|
| 57 |
+
"patch_size": 16,
|
| 58 |
+
"resample": 3,
|
| 59 |
+
"rescale_factor": 0.00392156862745098,
|
| 60 |
+
"return_attention_mask": true,
|
| 61 |
+
"sampling_rate": 16000,
|
| 62 |
+
"size": {
|
| 63 |
+
"longest_edge": 12845056,
|
| 64 |
+
"shortest_edge": 3136
|
| 65 |
+
},
|
| 66 |
+
"temporal_patch_size": 2
|
| 67 |
+
},
|
| 68 |
+
"processor_class": "Qwen3OmniMoeProcessor",
|
| 69 |
+
"video_processor": {
|
| 70 |
+
"dither": 0.0,
|
| 71 |
+
"do_convert_rgb": true,
|
| 72 |
+
"do_normalize": true,
|
| 73 |
+
"do_rescale": true,
|
| 74 |
+
"do_resize": true,
|
| 75 |
+
"do_sample_frames": false,
|
| 76 |
+
"feature_size": 128,
|
| 77 |
+
"hop_length": 160,
|
| 78 |
+
"image_mean": [
|
| 79 |
+
0.5,
|
| 80 |
+
0.5,
|
| 81 |
+
0.5
|
| 82 |
+
],
|
| 83 |
+
"image_std": [
|
| 84 |
+
0.5,
|
| 85 |
+
0.5,
|
| 86 |
+
0.5
|
| 87 |
+
],
|
| 88 |
+
"max_frames": 768,
|
| 89 |
+
"merge_size": 2,
|
| 90 |
+
"min_frames": 4,
|
| 91 |
+
"n_fft": 400,
|
| 92 |
+
"n_samples": 4800000,
|
| 93 |
+
"nb_max_frames": 30000,
|
| 94 |
+
"padding_side": "right",
|
| 95 |
+
"padding_value": 0.0,
|
| 96 |
+
"patch_size": 16,
|
| 97 |
+
"resample": 3,
|
| 98 |
+
"rescale_factor": 0.00392156862745098,
|
| 99 |
+
"return_attention_mask": true,
|
| 100 |
+
"return_metadata": false,
|
| 101 |
+
"sampling_rate": 16000,
|
| 102 |
+
"size": {
|
| 103 |
+
"longest_edge": 12845056,
|
| 104 |
+
"shortest_edge": 3136
|
| 105 |
+
},
|
| 106 |
+
"temporal_patch_size": 2,
|
| 107 |
+
"video_processor_type": "Qwen2VLVideoProcessor"
|
| 108 |
+
}
|
| 109 |
+
}
|
VA-Judger/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7d41145c408f3062e824f965be1c29854cd809e111cdf8170aa8b0bcd5d1fab
|
| 3 |
+
size 11424262
|
VA-Judger/tokenizer_config.json
ADDED
|
@@ -0,0 +1,316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<tool_response>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": false
|
| 188 |
+
},
|
| 189 |
+
"151666": {
|
| 190 |
+
"content": "</tool_response>",
|
| 191 |
+
"lstrip": false,
|
| 192 |
+
"normalized": false,
|
| 193 |
+
"rstrip": false,
|
| 194 |
+
"single_word": false,
|
| 195 |
+
"special": false
|
| 196 |
+
},
|
| 197 |
+
"151667": {
|
| 198 |
+
"content": "<think>",
|
| 199 |
+
"lstrip": false,
|
| 200 |
+
"normalized": false,
|
| 201 |
+
"rstrip": false,
|
| 202 |
+
"single_word": false,
|
| 203 |
+
"special": false
|
| 204 |
+
},
|
| 205 |
+
"151668": {
|
| 206 |
+
"content": "</think>",
|
| 207 |
+
"lstrip": false,
|
| 208 |
+
"normalized": false,
|
| 209 |
+
"rstrip": false,
|
| 210 |
+
"single_word": false,
|
| 211 |
+
"special": false
|
| 212 |
+
},
|
| 213 |
+
"151669": {
|
| 214 |
+
"content": "<|audio_start|>",
|
| 215 |
+
"lstrip": false,
|
| 216 |
+
"normalized": false,
|
| 217 |
+
"rstrip": false,
|
| 218 |
+
"single_word": false,
|
| 219 |
+
"special": true
|
| 220 |
+
},
|
| 221 |
+
"151670": {
|
| 222 |
+
"content": "<|audio_end|>",
|
| 223 |
+
"lstrip": false,
|
| 224 |
+
"normalized": false,
|
| 225 |
+
"rstrip": false,
|
| 226 |
+
"single_word": false,
|
| 227 |
+
"special": true
|
| 228 |
+
},
|
| 229 |
+
"151671": {
|
| 230 |
+
"content": "<tts_pad>",
|
| 231 |
+
"lstrip": false,
|
| 232 |
+
"normalized": false,
|
| 233 |
+
"rstrip": false,
|
| 234 |
+
"single_word": false,
|
| 235 |
+
"special": true
|
| 236 |
+
},
|
| 237 |
+
"151672": {
|
| 238 |
+
"content": "<tts_text_bos>",
|
| 239 |
+
"lstrip": false,
|
| 240 |
+
"normalized": false,
|
| 241 |
+
"rstrip": false,
|
| 242 |
+
"single_word": false,
|
| 243 |
+
"special": true
|
| 244 |
+
},
|
| 245 |
+
"151673": {
|
| 246 |
+
"content": "<tts_text_eod>",
|
| 247 |
+
"lstrip": false,
|
| 248 |
+
"normalized": false,
|
| 249 |
+
"rstrip": false,
|
| 250 |
+
"single_word": false,
|
| 251 |
+
"special": true
|
| 252 |
+
},
|
| 253 |
+
"151674": {
|
| 254 |
+
"content": "<tts_text_bos_single>",
|
| 255 |
+
"lstrip": false,
|
| 256 |
+
"normalized": false,
|
| 257 |
+
"rstrip": false,
|
| 258 |
+
"single_word": false,
|
| 259 |
+
"special": true
|
| 260 |
+
},
|
| 261 |
+
"151675": {
|
| 262 |
+
"content": "<|audio_pad|>",
|
| 263 |
+
"lstrip": false,
|
| 264 |
+
"normalized": false,
|
| 265 |
+
"rstrip": false,
|
| 266 |
+
"single_word": false,
|
| 267 |
+
"special": true
|
| 268 |
+
}
|
| 269 |
+
},
|
| 270 |
+
"additional_special_tokens": [
|
| 271 |
+
"<|im_start|>",
|
| 272 |
+
"<|im_end|>",
|
| 273 |
+
"<|object_ref_start|>",
|
| 274 |
+
"<|object_ref_end|>",
|
| 275 |
+
"<|box_start|>",
|
| 276 |
+
"<|box_end|>",
|
| 277 |
+
"<|quad_start|>",
|
| 278 |
+
"<|quad_end|>",
|
| 279 |
+
"<|vision_start|>",
|
| 280 |
+
"<|vision_end|>",
|
| 281 |
+
"<|vision_pad|>",
|
| 282 |
+
"<|image_pad|>",
|
| 283 |
+
"<|video_pad|>",
|
| 284 |
+
"<|audio_start|>",
|
| 285 |
+
"<|audio_end|>",
|
| 286 |
+
"<tts_pad>",
|
| 287 |
+
"<tts_text_bos>",
|
| 288 |
+
"<tts_text_bos_single>",
|
| 289 |
+
"<|audio_pad|>"
|
| 290 |
+
],
|
| 291 |
+
"extra_special_tokens": {
|
| 292 |
+
"image_token": "<|image_pad|>",
|
| 293 |
+
"audio_token": "<|audio_pad|>",
|
| 294 |
+
"video_token": "<|video_pad|>",
|
| 295 |
+
"vision_bos_token": "<|vision_start|>",
|
| 296 |
+
"vision_eos_token": "<|vision_end|>",
|
| 297 |
+
"audio_bos_token": "<|audio_start|>",
|
| 298 |
+
"audio_eos_token": "<|audio_end|>"
|
| 299 |
+
},
|
| 300 |
+
"bos_token": null,
|
| 301 |
+
"clean_up_tokenization_spaces": false,
|
| 302 |
+
"eos_token": "<|im_end|>",
|
| 303 |
+
"errors": "replace",
|
| 304 |
+
"model_max_length": 131072,
|
| 305 |
+
"pad_token": "<|endoftext|>",
|
| 306 |
+
"split_special_tokens": false,
|
| 307 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 308 |
+
"unk_token": null,
|
| 309 |
+
"image_token": "<|image_pad|>",
|
| 310 |
+
"audio_token": "<|audio_pad|>",
|
| 311 |
+
"video_token": "<|video_pad|>",
|
| 312 |
+
"vision_bos_token": "<|vision_start|>",
|
| 313 |
+
"vision_eos_token": "<|vision_end|>",
|
| 314 |
+
"audio_bos_token": "<|audio_start|>",
|
| 315 |
+
"audio_eos_token": "<|audio_end|>"
|
| 316 |
+
}
|
VA-Judger/trainer_state.json
ADDED
|
@@ -0,0 +1,1962 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 17.395604395604394,
|
| 6 |
+
"eval_steps": 100.0,
|
| 7 |
+
"global_step": 1200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.014652014652014652,
|
| 14 |
+
"grad_norm": 8.054882049560547,
|
| 15 |
+
"learning_rate": 5.0000000000000004e-08,
|
| 16 |
+
"loss": 0.9730892181396484,
|
| 17 |
+
"step": 1,
|
| 18 |
+
"token_acc": 0.755182027556669
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.07326007326007326,
|
| 22 |
+
"grad_norm": 8.133082389831543,
|
| 23 |
+
"learning_rate": 2.5000000000000004e-07,
|
| 24 |
+
"loss": 0.974563479423523,
|
| 25 |
+
"step": 5,
|
| 26 |
+
"token_acc": 0.7555487882248446
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.14652014652014653,
|
| 30 |
+
"grad_norm": 7.104198932647705,
|
| 31 |
+
"learning_rate": 5.000000000000001e-07,
|
| 32 |
+
"loss": 0.9796989440917969,
|
| 33 |
+
"step": 10,
|
| 34 |
+
"token_acc": 0.7552294244705977
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.21978021978021978,
|
| 38 |
+
"grad_norm": 4.173017501831055,
|
| 39 |
+
"learning_rate": 7.5e-07,
|
| 40 |
+
"loss": 0.9597454071044922,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"token_acc": 0.7582719591133184
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.29304029304029305,
|
| 46 |
+
"grad_norm": 3.174144744873047,
|
| 47 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 48 |
+
"loss": 0.905433464050293,
|
| 49 |
+
"step": 20,
|
| 50 |
+
"token_acc": 0.7677386934673367
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.3663003663003663,
|
| 54 |
+
"grad_norm": 2.7379891872406006,
|
| 55 |
+
"learning_rate": 1.25e-06,
|
| 56 |
+
"loss": 0.8424328804016114,
|
| 57 |
+
"step": 25,
|
| 58 |
+
"token_acc": 0.7794254501314991
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.43956043956043955,
|
| 62 |
+
"grad_norm": 2.221156597137451,
|
| 63 |
+
"learning_rate": 1.5e-06,
|
| 64 |
+
"loss": 0.7975196838378906,
|
| 65 |
+
"step": 30,
|
| 66 |
+
"token_acc": 0.7875305544575223
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5128205128205128,
|
| 70 |
+
"grad_norm": 1.708078384399414,
|
| 71 |
+
"learning_rate": 1.75e-06,
|
| 72 |
+
"loss": 0.7645464897155761,
|
| 73 |
+
"step": 35,
|
| 74 |
+
"token_acc": 0.7928033412196417
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.5860805860805861,
|
| 78 |
+
"grad_norm": 1.729050636291504,
|
| 79 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 80 |
+
"loss": 0.7186562538146972,
|
| 81 |
+
"step": 40,
|
| 82 |
+
"token_acc": 0.8009626994350602
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.6593406593406593,
|
| 86 |
+
"grad_norm": 1.3434247970581055,
|
| 87 |
+
"learning_rate": 2.25e-06,
|
| 88 |
+
"loss": 0.6939143180847168,
|
| 89 |
+
"step": 45,
|
| 90 |
+
"token_acc": 0.806187490700265
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.7326007326007326,
|
| 94 |
+
"grad_norm": 1.3713105916976929,
|
| 95 |
+
"learning_rate": 2.5e-06,
|
| 96 |
+
"loss": 0.6725265502929687,
|
| 97 |
+
"step": 50,
|
| 98 |
+
"token_acc": 0.8099984695809068
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.8058608058608059,
|
| 102 |
+
"grad_norm": 1.4401501417160034,
|
| 103 |
+
"learning_rate": 2.7500000000000004e-06,
|
| 104 |
+
"loss": 0.6566809177398681,
|
| 105 |
+
"step": 55,
|
| 106 |
+
"token_acc": 0.8134286266799884
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.8791208791208791,
|
| 110 |
+
"grad_norm": 1.3608494997024536,
|
| 111 |
+
"learning_rate": 3e-06,
|
| 112 |
+
"loss": 0.6351348876953125,
|
| 113 |
+
"step": 60,
|
| 114 |
+
"token_acc": 0.8176798262114832
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.9523809523809523,
|
| 118 |
+
"grad_norm": 1.3621470928192139,
|
| 119 |
+
"learning_rate": 3.2500000000000002e-06,
|
| 120 |
+
"loss": 0.6236963748931885,
|
| 121 |
+
"step": 65,
|
| 122 |
+
"token_acc": 0.8199955690721235
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.0146520146520146,
|
| 126 |
+
"grad_norm": 1.122348666191101,
|
| 127 |
+
"learning_rate": 3.5e-06,
|
| 128 |
+
"loss": 0.5974433422088623,
|
| 129 |
+
"step": 70,
|
| 130 |
+
"token_acc": 0.8218195920883842
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.0879120879120878,
|
| 134 |
+
"grad_norm": 1.0651134252548218,
|
| 135 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 136 |
+
"loss": 0.599330997467041,
|
| 137 |
+
"step": 75,
|
| 138 |
+
"token_acc": 0.8252219321148825
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.1611721611721613,
|
| 142 |
+
"grad_norm": 1.0729129314422607,
|
| 143 |
+
"learning_rate": 4.000000000000001e-06,
|
| 144 |
+
"loss": 0.5846376419067383,
|
| 145 |
+
"step": 80,
|
| 146 |
+
"token_acc": 0.828906085747093
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.2344322344322345,
|
| 150 |
+
"grad_norm": 1.3562551736831665,
|
| 151 |
+
"learning_rate": 4.25e-06,
|
| 152 |
+
"loss": 0.5784817695617676,
|
| 153 |
+
"step": 85,
|
| 154 |
+
"token_acc": 0.8290547980859712
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.3076923076923077,
|
| 158 |
+
"grad_norm": 2.5216593742370605,
|
| 159 |
+
"learning_rate": 4.5e-06,
|
| 160 |
+
"loss": 0.5747579574584961,
|
| 161 |
+
"step": 90,
|
| 162 |
+
"token_acc": 0.8305322695494562
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.380952380952381,
|
| 166 |
+
"grad_norm": 1.4898324012756348,
|
| 167 |
+
"learning_rate": 4.75e-06,
|
| 168 |
+
"loss": 0.5692717075347901,
|
| 169 |
+
"step": 95,
|
| 170 |
+
"token_acc": 0.8308044788000502
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.4542124542124542,
|
| 174 |
+
"grad_norm": 1.333472490310669,
|
| 175 |
+
"learning_rate": 5e-06,
|
| 176 |
+
"loss": 0.5596965312957763,
|
| 177 |
+
"step": 100,
|
| 178 |
+
"token_acc": 0.8323691980915754
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.5274725274725274,
|
| 182 |
+
"grad_norm": 10.963144302368164,
|
| 183 |
+
"learning_rate": 4.999914564160437e-06,
|
| 184 |
+
"loss": 0.548871374130249,
|
| 185 |
+
"step": 105,
|
| 186 |
+
"token_acc": 0.8353487211277584
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.6007326007326008,
|
| 190 |
+
"grad_norm": 1.1092952489852905,
|
| 191 |
+
"learning_rate": 4.999658262481173e-06,
|
| 192 |
+
"loss": 0.5265289783477783,
|
| 193 |
+
"step": 110,
|
| 194 |
+
"token_acc": 0.8404660153428328
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.673992673992674,
|
| 198 |
+
"grad_norm": 1.171152949333191,
|
| 199 |
+
"learning_rate": 4.999231112480088e-06,
|
| 200 |
+
"loss": 0.5289466857910157,
|
| 201 |
+
"step": 115,
|
| 202 |
+
"token_acc": 0.8396460589312683
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 1.7472527472527473,
|
| 206 |
+
"grad_norm": 6.100719928741455,
|
| 207 |
+
"learning_rate": 4.998633143352315e-06,
|
| 208 |
+
"loss": 0.5207566261291504,
|
| 209 |
+
"step": 120,
|
| 210 |
+
"token_acc": 0.8416935558777325
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.8205128205128205,
|
| 214 |
+
"grad_norm": 2.674548625946045,
|
| 215 |
+
"learning_rate": 4.997864395968252e-06,
|
| 216 |
+
"loss": 0.5166749954223633,
|
| 217 |
+
"step": 125,
|
| 218 |
+
"token_acc": 0.8429160713926394
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.8937728937728937,
|
| 222 |
+
"grad_norm": 1.5767267942428589,
|
| 223 |
+
"learning_rate": 4.9969249228707625e-06,
|
| 224 |
+
"loss": 0.5092305183410645,
|
| 225 |
+
"step": 130,
|
| 226 |
+
"token_acc": 0.844041897620484
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.9670329670329672,
|
| 230 |
+
"grad_norm": 1.3257472515106201,
|
| 231 |
+
"learning_rate": 4.995814788271582e-06,
|
| 232 |
+
"loss": 0.5064918041229248,
|
| 233 |
+
"step": 135,
|
| 234 |
+
"token_acc": 0.8438894312616487
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.029304029304029,
|
| 238 |
+
"grad_norm": 1.4548439979553223,
|
| 239 |
+
"learning_rate": 4.994534068046936e-06,
|
| 240 |
+
"loss": 0.48206567764282227,
|
| 241 |
+
"step": 140,
|
| 242 |
+
"token_acc": 0.8433812761635217
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1025641025641026,
|
| 246 |
+
"grad_norm": 1.3242915868759155,
|
| 247 |
+
"learning_rate": 4.993082849732353e-06,
|
| 248 |
+
"loss": 0.49968156814575193,
|
| 249 |
+
"step": 145,
|
| 250 |
+
"token_acc": 0.8467234610805419
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 2.1758241758241756,
|
| 254 |
+
"grad_norm": 1.3747650384902954,
|
| 255 |
+
"learning_rate": 4.991461232516675e-06,
|
| 256 |
+
"loss": 0.48995189666748046,
|
| 257 |
+
"step": 150,
|
| 258 |
+
"token_acc": 0.8483358013533089
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.249084249084249,
|
| 262 |
+
"grad_norm": 1.2035170793533325,
|
| 263 |
+
"learning_rate": 4.989669327235285e-06,
|
| 264 |
+
"loss": 0.48351707458496096,
|
| 265 |
+
"step": 155,
|
| 266 |
+
"token_acc": 0.85017818093464
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.3223443223443225,
|
| 270 |
+
"grad_norm": 1.2826895713806152,
|
| 271 |
+
"learning_rate": 4.987707256362529e-06,
|
| 272 |
+
"loss": 0.4865525245666504,
|
| 273 |
+
"step": 160,
|
| 274 |
+
"token_acc": 0.8483052731572246
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.3956043956043955,
|
| 278 |
+
"grad_norm": 1.269644856452942,
|
| 279 |
+
"learning_rate": 4.9855751540033446e-06,
|
| 280 |
+
"loss": 0.48340187072753904,
|
| 281 |
+
"step": 165,
|
| 282 |
+
"token_acc": 0.8504842958459979
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.468864468864469,
|
| 286 |
+
"grad_norm": 1.6188918352127075,
|
| 287 |
+
"learning_rate": 4.983273165884096e-06,
|
| 288 |
+
"loss": 0.4890533447265625,
|
| 289 |
+
"step": 170,
|
| 290 |
+
"token_acc": 0.8496313114062406
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.542124542124542,
|
| 294 |
+
"grad_norm": 1.2251207828521729,
|
| 295 |
+
"learning_rate": 4.9808014493426124e-06,
|
| 296 |
+
"loss": 0.4855842590332031,
|
| 297 |
+
"step": 175,
|
| 298 |
+
"token_acc": 0.8491935320664602
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.6153846153846154,
|
| 302 |
+
"grad_norm": 1.5653579235076904,
|
| 303 |
+
"learning_rate": 4.978160173317439e-06,
|
| 304 |
+
"loss": 0.48499269485473634,
|
| 305 |
+
"step": 180,
|
| 306 |
+
"token_acc": 0.8498993720817903
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 2.688644688644689,
|
| 310 |
+
"grad_norm": 1.005551815032959,
|
| 311 |
+
"learning_rate": 4.97534951833628e-06,
|
| 312 |
+
"loss": 0.48335814476013184,
|
| 313 |
+
"step": 185,
|
| 314 |
+
"token_acc": 0.8504971734098918
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.761904761904762,
|
| 318 |
+
"grad_norm": 1.1992868185043335,
|
| 319 |
+
"learning_rate": 4.972369676503672e-06,
|
| 320 |
+
"loss": 0.47909183502197267,
|
| 321 |
+
"step": 190,
|
| 322 |
+
"token_acc": 0.8512129881044738
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 2.8351648351648353,
|
| 326 |
+
"grad_norm": 1.490403175354004,
|
| 327 |
+
"learning_rate": 4.9692208514878445e-06,
|
| 328 |
+
"loss": 0.48547868728637694,
|
| 329 |
+
"step": 195,
|
| 330 |
+
"token_acc": 0.8496412539776645
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.9084249084249083,
|
| 334 |
+
"grad_norm": 1.525592565536499,
|
| 335 |
+
"learning_rate": 4.965903258506806e-06,
|
| 336 |
+
"loss": 0.48122386932373046,
|
| 337 |
+
"step": 200,
|
| 338 |
+
"token_acc": 0.8510984540276648
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.9816849816849818,
|
| 342 |
+
"grad_norm": 1.3033020496368408,
|
| 343 |
+
"learning_rate": 4.96241712431363e-06,
|
| 344 |
+
"loss": 0.4055511474609375,
|
| 345 |
+
"step": 205,
|
| 346 |
+
"token_acc": 0.8697563663715604
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.043956043956044,
|
| 350 |
+
"grad_norm": 1.7497599124908447,
|
| 351 |
+
"learning_rate": 4.9587626871809564e-06,
|
| 352 |
+
"loss": 0.37597928047180174,
|
| 353 |
+
"step": 210,
|
| 354 |
+
"token_acc": 0.8658160145939685
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 3.1172161172161172,
|
| 358 |
+
"grad_norm": 1.5791239738464355,
|
| 359 |
+
"learning_rate": 4.95494019688471e-06,
|
| 360 |
+
"loss": 0.434743595123291,
|
| 361 |
+
"step": 215,
|
| 362 |
+
"token_acc": 0.8626484680114346
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.1904761904761907,
|
| 366 |
+
"grad_norm": 2.850130796432495,
|
| 367 |
+
"learning_rate": 4.950949914687024e-06,
|
| 368 |
+
"loss": 0.4336540222167969,
|
| 369 |
+
"step": 220,
|
| 370 |
+
"token_acc": 0.8629487550665894
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 3.2637362637362637,
|
| 374 |
+
"grad_norm": 2.0345170497894287,
|
| 375 |
+
"learning_rate": 4.9467921133183864e-06,
|
| 376 |
+
"loss": 0.430650520324707,
|
| 377 |
+
"step": 225,
|
| 378 |
+
"token_acc": 0.8644715738019915
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 3.336996336996337,
|
| 382 |
+
"grad_norm": 1.5538909435272217,
|
| 383 |
+
"learning_rate": 4.942467076958999e-06,
|
| 384 |
+
"loss": 0.4266399383544922,
|
| 385 |
+
"step": 230,
|
| 386 |
+
"token_acc": 0.8640332223247602
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.41025641025641,
|
| 390 |
+
"grad_norm": 1.5413146018981934,
|
| 391 |
+
"learning_rate": 4.93797510121935e-06,
|
| 392 |
+
"loss": 0.4310781002044678,
|
| 393 |
+
"step": 235,
|
| 394 |
+
"token_acc": 0.8633472402708433
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 3.4835164835164836,
|
| 398 |
+
"grad_norm": 1.183243751525879,
|
| 399 |
+
"learning_rate": 4.933316493120015e-06,
|
| 400 |
+
"loss": 0.4210788726806641,
|
| 401 |
+
"step": 240,
|
| 402 |
+
"token_acc": 0.865890725175681
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.5567765567765566,
|
| 406 |
+
"grad_norm": 1.0250470638275146,
|
| 407 |
+
"learning_rate": 4.928491571070669e-06,
|
| 408 |
+
"loss": 0.4344191551208496,
|
| 409 |
+
"step": 245,
|
| 410 |
+
"token_acc": 0.8623012817152971
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 3.63003663003663,
|
| 414 |
+
"grad_norm": 3.5557236671447754,
|
| 415 |
+
"learning_rate": 4.923500664848327e-06,
|
| 416 |
+
"loss": 0.4351386547088623,
|
| 417 |
+
"step": 250,
|
| 418 |
+
"token_acc": 0.8621746651112273
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 3.7032967032967035,
|
| 422 |
+
"grad_norm": 2.2677974700927734,
|
| 423 |
+
"learning_rate": 4.918344115574797e-06,
|
| 424 |
+
"loss": 0.43351593017578127,
|
| 425 |
+
"step": 255,
|
| 426 |
+
"token_acc": 0.8624701381714877
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 3.7765567765567765,
|
| 430 |
+
"grad_norm": 1.3171656131744385,
|
| 431 |
+
"learning_rate": 4.913022275693372e-06,
|
| 432 |
+
"loss": 0.42925281524658204,
|
| 433 |
+
"step": 260,
|
| 434 |
+
"token_acc": 0.8637055603086428
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.84981684981685,
|
| 438 |
+
"grad_norm": 4.644964218139648,
|
| 439 |
+
"learning_rate": 4.90753550894474e-06,
|
| 440 |
+
"loss": 0.42578854560852053,
|
| 441 |
+
"step": 265,
|
| 442 |
+
"token_acc": 0.8650916586371342
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.9230769230769234,
|
| 446 |
+
"grad_norm": 1.2326524257659912,
|
| 447 |
+
"learning_rate": 4.901884190342121e-06,
|
| 448 |
+
"loss": 0.4164420127868652,
|
| 449 |
+
"step": 270,
|
| 450 |
+
"token_acc": 0.8665119613155625
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 3.9963369963369964,
|
| 454 |
+
"grad_norm": 1.141901969909668,
|
| 455 |
+
"learning_rate": 4.896068706145632e-06,
|
| 456 |
+
"loss": 0.43213911056518556,
|
| 457 |
+
"step": 275,
|
| 458 |
+
"token_acc": 0.862691240362133
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.058608058608058,
|
| 462 |
+
"grad_norm": 1.1900097131729126,
|
| 463 |
+
"learning_rate": 4.890089453835894e-06,
|
| 464 |
+
"loss": 0.37670063972473145,
|
| 465 |
+
"step": 280,
|
| 466 |
+
"token_acc": 0.879840770836551
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.131868131868132,
|
| 470 |
+
"grad_norm": 1.2517964839935303,
|
| 471 |
+
"learning_rate": 4.883946842086861e-06,
|
| 472 |
+
"loss": 0.36878142356872556,
|
| 473 |
+
"step": 285,
|
| 474 |
+
"token_acc": 0.8808383281903808
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.205128205128205,
|
| 478 |
+
"grad_norm": 2.362907648086548,
|
| 479 |
+
"learning_rate": 4.8776412907378845e-06,
|
| 480 |
+
"loss": 0.36941800117492674,
|
| 481 |
+
"step": 290,
|
| 482 |
+
"token_acc": 0.8811055235623327
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.278388278388278,
|
| 486 |
+
"grad_norm": 1.6611422300338745,
|
| 487 |
+
"learning_rate": 4.871173230765024e-06,
|
| 488 |
+
"loss": 0.3731838226318359,
|
| 489 |
+
"step": 295,
|
| 490 |
+
"token_acc": 0.8791506164102647
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 4.351648351648351,
|
| 494 |
+
"grad_norm": 1.0864579677581787,
|
| 495 |
+
"learning_rate": 4.864543104251587e-06,
|
| 496 |
+
"loss": 0.3697177886962891,
|
| 497 |
+
"step": 300,
|
| 498 |
+
"token_acc": 0.8797446986382855
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 4.424908424908425,
|
| 502 |
+
"grad_norm": 1.6939606666564941,
|
| 503 |
+
"learning_rate": 4.857751364357913e-06,
|
| 504 |
+
"loss": 0.37488179206848143,
|
| 505 |
+
"step": 305,
|
| 506 |
+
"token_acc": 0.8791811413388549
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.498168498168498,
|
| 510 |
+
"grad_norm": 1.0882139205932617,
|
| 511 |
+
"learning_rate": 4.850798475290403e-06,
|
| 512 |
+
"loss": 0.35840139389038084,
|
| 513 |
+
"step": 310,
|
| 514 |
+
"token_acc": 0.883983676095129
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.571428571428571,
|
| 518 |
+
"grad_norm": 1.1809661388397217,
|
| 519 |
+
"learning_rate": 4.843684912269789e-06,
|
| 520 |
+
"loss": 0.34653630256652834,
|
| 521 |
+
"step": 315,
|
| 522 |
+
"token_acc": 0.8870768460494864
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 4.644688644688645,
|
| 526 |
+
"grad_norm": 1.629836082458496,
|
| 527 |
+
"learning_rate": 4.836411161498653e-06,
|
| 528 |
+
"loss": 0.33505845069885254,
|
| 529 |
+
"step": 320,
|
| 530 |
+
"token_acc": 0.8900206447259553
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.717948717948718,
|
| 534 |
+
"grad_norm": 1.0773942470550537,
|
| 535 |
+
"learning_rate": 4.828977720128198e-06,
|
| 536 |
+
"loss": 0.34183175563812257,
|
| 537 |
+
"step": 325,
|
| 538 |
+
"token_acc": 0.8875660151960646
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.791208791208791,
|
| 542 |
+
"grad_norm": 1.1171730756759644,
|
| 543 |
+
"learning_rate": 4.821385096224268e-06,
|
| 544 |
+
"loss": 0.3330850601196289,
|
| 545 |
+
"step": 330,
|
| 546 |
+
"token_acc": 0.890400433513155
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 4.864468864468865,
|
| 550 |
+
"grad_norm": 1.38460373878479,
|
| 551 |
+
"learning_rate": 4.8136338087326214e-06,
|
| 552 |
+
"loss": 0.3419033050537109,
|
| 553 |
+
"step": 335,
|
| 554 |
+
"token_acc": 0.8880274585930686
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 4.937728937728938,
|
| 558 |
+
"grad_norm": 1.623136281967163,
|
| 559 |
+
"learning_rate": 4.8057243874434625e-06,
|
| 560 |
+
"loss": 0.3189419746398926,
|
| 561 |
+
"step": 340,
|
| 562 |
+
"token_acc": 0.89473237832533
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.0,
|
| 566 |
+
"grad_norm": 1.974323034286499,
|
| 567 |
+
"learning_rate": 4.797657372955228e-06,
|
| 568 |
+
"loss": 0.23971011638641357,
|
| 569 |
+
"step": 345,
|
| 570 |
+
"token_acc": 0.9065612695441807
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 5.073260073260073,
|
| 574 |
+
"grad_norm": 1.1910440921783447,
|
| 575 |
+
"learning_rate": 4.789433316637644e-06,
|
| 576 |
+
"loss": 0.31196768283843995,
|
| 577 |
+
"step": 350,
|
| 578 |
+
"token_acc": 0.8987927152024427
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 5.146520146520147,
|
| 582 |
+
"grad_norm": 1.3556087017059326,
|
| 583 |
+
"learning_rate": 4.7810527805940344e-06,
|
| 584 |
+
"loss": 0.3126432180404663,
|
| 585 |
+
"step": 355,
|
| 586 |
+
"token_acc": 0.8989477503628447
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 5.21978021978022,
|
| 590 |
+
"grad_norm": 1.234954833984375,
|
| 591 |
+
"learning_rate": 4.772516337622907e-06,
|
| 592 |
+
"loss": 0.3166388988494873,
|
| 593 |
+
"step": 360,
|
| 594 |
+
"token_acc": 0.8971954330575381
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 5.293040293040293,
|
| 598 |
+
"grad_norm": 1.0847809314727783,
|
| 599 |
+
"learning_rate": 4.763824571178798e-06,
|
| 600 |
+
"loss": 0.316084885597229,
|
| 601 |
+
"step": 365,
|
| 602 |
+
"token_acc": 0.8972399116578633
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 5.366300366300366,
|
| 606 |
+
"grad_norm": 1.5089348554611206,
|
| 607 |
+
"learning_rate": 4.754978075332398e-06,
|
| 608 |
+
"loss": 0.31339383125305176,
|
| 609 |
+
"step": 370,
|
| 610 |
+
"token_acc": 0.8984290793262657
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 5.43956043956044,
|
| 614 |
+
"grad_norm": 1.0931137800216675,
|
| 615 |
+
"learning_rate": 4.745977454729947e-06,
|
| 616 |
+
"loss": 0.3083870649337769,
|
| 617 |
+
"step": 375,
|
| 618 |
+
"token_acc": 0.8992366350745247
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.512820512820513,
|
| 622 |
+
"grad_norm": 1.2304120063781738,
|
| 623 |
+
"learning_rate": 4.736823324551909e-06,
|
| 624 |
+
"loss": 0.31461353302001954,
|
| 625 |
+
"step": 380,
|
| 626 |
+
"token_acc": 0.8981053020239276
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 5.586080586080586,
|
| 630 |
+
"grad_norm": 1.027147650718689,
|
| 631 |
+
"learning_rate": 4.72751631047092e-06,
|
| 632 |
+
"loss": 0.31214213371276855,
|
| 633 |
+
"step": 385,
|
| 634 |
+
"token_acc": 0.8970606028829323
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 5.65934065934066,
|
| 638 |
+
"grad_norm": 1.1977190971374512,
|
| 639 |
+
"learning_rate": 4.71805704860903e-06,
|
| 640 |
+
"loss": 0.3054037094116211,
|
| 641 |
+
"step": 390,
|
| 642 |
+
"token_acc": 0.8994789933697442
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 5.732600732600733,
|
| 646 |
+
"grad_norm": 1.1509610414505005,
|
| 647 |
+
"learning_rate": 4.708446185494222e-06,
|
| 648 |
+
"loss": 0.3100709915161133,
|
| 649 |
+
"step": 395,
|
| 650 |
+
"token_acc": 0.8986802153758957
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 5.805860805860806,
|
| 654 |
+
"grad_norm": 1.0790588855743408,
|
| 655 |
+
"learning_rate": 4.698684378016223e-06,
|
| 656 |
+
"loss": 0.3133284568786621,
|
| 657 |
+
"step": 400,
|
| 658 |
+
"token_acc": 0.8975068693234776
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 5.8791208791208796,
|
| 662 |
+
"grad_norm": 2.208139181137085,
|
| 663 |
+
"learning_rate": 4.688772293381608e-06,
|
| 664 |
+
"loss": 0.22988929748535156,
|
| 665 |
+
"step": 405,
|
| 666 |
+
"token_acc": 0.9242264266961169
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 5.9523809523809526,
|
| 670 |
+
"grad_norm": 1.3288463354110718,
|
| 671 |
+
"learning_rate": 4.678710609068193e-06,
|
| 672 |
+
"loss": 0.19912091493606568,
|
| 673 |
+
"step": 410,
|
| 674 |
+
"token_acc": 0.9330849328311616
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.014652014652015,
|
| 678 |
+
"grad_norm": 3.163670778274536,
|
| 679 |
+
"learning_rate": 4.668500012778738e-06,
|
| 680 |
+
"loss": 0.1649259328842163,
|
| 681 |
+
"step": 415,
|
| 682 |
+
"token_acc": 0.9344287553242777
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 6.087912087912088,
|
| 686 |
+
"grad_norm": 1.6831578016281128,
|
| 687 |
+
"learning_rate": 4.658141202393935e-06,
|
| 688 |
+
"loss": 0.254819393157959,
|
| 689 |
+
"step": 420,
|
| 690 |
+
"token_acc": 0.9172843213806635
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 6.1611721611721615,
|
| 694 |
+
"grad_norm": 1.4219450950622559,
|
| 695 |
+
"learning_rate": 4.647634885924713e-06,
|
| 696 |
+
"loss": 0.24539401531219482,
|
| 697 |
+
"step": 425,
|
| 698 |
+
"token_acc": 0.9202387653531062
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 6.2344322344322345,
|
| 702 |
+
"grad_norm": 3.1420774459838867,
|
| 703 |
+
"learning_rate": 4.636981781463848e-06,
|
| 704 |
+
"loss": 0.2495945453643799,
|
| 705 |
+
"step": 430,
|
| 706 |
+
"token_acc": 0.9186168242361771
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.3076923076923075,
|
| 710 |
+
"grad_norm": 1.2774112224578857,
|
| 711 |
+
"learning_rate": 4.626182617136877e-06,
|
| 712 |
+
"loss": 0.24089446067810058,
|
| 713 |
+
"step": 435,
|
| 714 |
+
"token_acc": 0.9207879973485522
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 6.380952380952381,
|
| 718 |
+
"grad_norm": 1.296355128288269,
|
| 719 |
+
"learning_rate": 4.615238131052339e-06,
|
| 720 |
+
"loss": 0.2517597198486328,
|
| 721 |
+
"step": 440,
|
| 722 |
+
"token_acc": 0.9169400229982448
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 6.454212454212454,
|
| 726 |
+
"grad_norm": 1.2661314010620117,
|
| 727 |
+
"learning_rate": 4.604149071251318e-06,
|
| 728 |
+
"loss": 0.25311870574951173,
|
| 729 |
+
"step": 445,
|
| 730 |
+
"token_acc": 0.9164257216236734
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 6.527472527472527,
|
| 734 |
+
"grad_norm": 1.229150414466858,
|
| 735 |
+
"learning_rate": 4.592916195656322e-06,
|
| 736 |
+
"loss": 0.25477099418640137,
|
| 737 |
+
"step": 450,
|
| 738 |
+
"token_acc": 0.9167566609177802
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 6.6007326007326,
|
| 742 |
+
"grad_norm": 1.1014615297317505,
|
| 743 |
+
"learning_rate": 4.581540272019476e-06,
|
| 744 |
+
"loss": 0.2602519989013672,
|
| 745 |
+
"step": 455,
|
| 746 |
+
"token_acc": 0.9153924075056679
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 6.673992673992674,
|
| 750 |
+
"grad_norm": 1.1742384433746338,
|
| 751 |
+
"learning_rate": 4.570022077870051e-06,
|
| 752 |
+
"loss": 0.25152149200439455,
|
| 753 |
+
"step": 460,
|
| 754 |
+
"token_acc": 0.9173155044008124
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 6.747252747252747,
|
| 758 |
+
"grad_norm": 1.1646785736083984,
|
| 759 |
+
"learning_rate": 4.558362400461315e-06,
|
| 760 |
+
"loss": 0.25430617332458494,
|
| 761 |
+
"step": 465,
|
| 762 |
+
"token_acc": 0.9158255725128805
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.82051282051282,
|
| 766 |
+
"grad_norm": 1.33492910861969,
|
| 767 |
+
"learning_rate": 4.546562036716732e-06,
|
| 768 |
+
"loss": 0.2558555603027344,
|
| 769 |
+
"step": 470,
|
| 770 |
+
"token_acc": 0.9154095240399543
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 6.893772893772894,
|
| 774 |
+
"grad_norm": 1.1138622760772705,
|
| 775 |
+
"learning_rate": 4.534621793175488e-06,
|
| 776 |
+
"loss": 0.2574252367019653,
|
| 777 |
+
"step": 475,
|
| 778 |
+
"token_acc": 0.9146798880809435
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 6.967032967032967,
|
| 782 |
+
"grad_norm": 1.2710987329483032,
|
| 783 |
+
"learning_rate": 4.522542485937369e-06,
|
| 784 |
+
"loss": 0.25526740550994875,
|
| 785 |
+
"step": 480,
|
| 786 |
+
"token_acc": 0.9157952713486828
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 7.029304029304029,
|
| 790 |
+
"grad_norm": 1.2122446298599243,
|
| 791 |
+
"learning_rate": 4.51032494060698e-06,
|
| 792 |
+
"loss": 0.22854809761047362,
|
| 793 |
+
"step": 485,
|
| 794 |
+
"token_acc": 0.9268327840821541
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.102564102564102,
|
| 798 |
+
"grad_norm": 1.1829161643981934,
|
| 799 |
+
"learning_rate": 4.497969992237312e-06,
|
| 800 |
+
"loss": 0.1869415521621704,
|
| 801 |
+
"step": 490,
|
| 802 |
+
"token_acc": 0.9378105128692299
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 7.175824175824176,
|
| 806 |
+
"grad_norm": 1.12260901927948,
|
| 807 |
+
"learning_rate": 4.485478485272678e-06,
|
| 808 |
+
"loss": 0.1821830153465271,
|
| 809 |
+
"step": 495,
|
| 810 |
+
"token_acc": 0.939871717779114
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 7.249084249084249,
|
| 814 |
+
"grad_norm": 1.1085420846939087,
|
| 815 |
+
"learning_rate": 4.472851273490985e-06,
|
| 816 |
+
"loss": 0.1860897183418274,
|
| 817 |
+
"step": 500,
|
| 818 |
+
"token_acc": 0.9374316333326606
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 7.322344322344322,
|
| 822 |
+
"grad_norm": 1.3176734447479248,
|
| 823 |
+
"learning_rate": 4.460089219945383e-06,
|
| 824 |
+
"loss": 0.2161034107208252,
|
| 825 |
+
"step": 505,
|
| 826 |
+
"token_acc": 0.9288069780301837
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 7.395604395604396,
|
| 830 |
+
"grad_norm": 1.2892754077911377,
|
| 831 |
+
"learning_rate": 4.4471931969052816e-06,
|
| 832 |
+
"loss": 0.1974965214729309,
|
| 833 |
+
"step": 510,
|
| 834 |
+
"token_acc": 0.9343932120310118
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 7.468864468864469,
|
| 838 |
+
"grad_norm": 1.2758907079696655,
|
| 839 |
+
"learning_rate": 4.434164085796724e-06,
|
| 840 |
+
"loss": 0.17482783794403076,
|
| 841 |
+
"step": 515,
|
| 842 |
+
"token_acc": 0.9417765463024865
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 7.542124542124542,
|
| 846 |
+
"grad_norm": 1.14219331741333,
|
| 847 |
+
"learning_rate": 4.421002777142148e-06,
|
| 848 |
+
"loss": 0.15864505767822265,
|
| 849 |
+
"step": 520,
|
| 850 |
+
"token_acc": 0.9470915929078415
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.615384615384615,
|
| 854 |
+
"grad_norm": 1.235621452331543,
|
| 855 |
+
"learning_rate": 4.407710170499517e-06,
|
| 856 |
+
"loss": 0.15338094234466554,
|
| 857 |
+
"step": 525,
|
| 858 |
+
"token_acc": 0.9486054880791723
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 7.688644688644689,
|
| 862 |
+
"grad_norm": 1.1506952047348022,
|
| 863 |
+
"learning_rate": 4.394287174400838e-06,
|
| 864 |
+
"loss": 0.16547095775604248,
|
| 865 |
+
"step": 530,
|
| 866 |
+
"token_acc": 0.9441399772753735
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 7.761904761904762,
|
| 870 |
+
"grad_norm": 1.1788743734359741,
|
| 871 |
+
"learning_rate": 4.380734706290063e-06,
|
| 872 |
+
"loss": 0.15452332496643068,
|
| 873 |
+
"step": 535,
|
| 874 |
+
"token_acc": 0.9481438022213681
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 7.835164835164835,
|
| 878 |
+
"grad_norm": 1.3817641735076904,
|
| 879 |
+
"learning_rate": 4.3670536924603855e-06,
|
| 880 |
+
"loss": 0.14616944789886474,
|
| 881 |
+
"step": 540,
|
| 882 |
+
"token_acc": 0.9509172297365357
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 7.908424908424909,
|
| 886 |
+
"grad_norm": 1.4607346057891846,
|
| 887 |
+
"learning_rate": 4.353245067990928e-06,
|
| 888 |
+
"loss": 0.12408863306045533,
|
| 889 |
+
"step": 545,
|
| 890 |
+
"token_acc": 0.9584563005785027
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 7.981684981684982,
|
| 894 |
+
"grad_norm": 1.0356935262680054,
|
| 895 |
+
"learning_rate": 4.33930977668283e-06,
|
| 896 |
+
"loss": 0.10001213550567627,
|
| 897 |
+
"step": 550,
|
| 898 |
+
"token_acc": 0.9668478064124274
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 8.043956043956044,
|
| 902 |
+
"grad_norm": 1.5036234855651855,
|
| 903 |
+
"learning_rate": 4.325248770994741e-06,
|
| 904 |
+
"loss": 0.11044188737869262,
|
| 905 |
+
"step": 555,
|
| 906 |
+
"token_acc": 0.9574441822113818
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 8.117216117216117,
|
| 910 |
+
"grad_norm": 1.2112280130386353,
|
| 911 |
+
"learning_rate": 4.311063011977723e-06,
|
| 912 |
+
"loss": 0.1332005262374878,
|
| 913 |
+
"step": 560,
|
| 914 |
+
"token_acc": 0.9566093714727982
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 8.19047619047619,
|
| 918 |
+
"grad_norm": 1.122559905052185,
|
| 919 |
+
"learning_rate": 4.29675346920956e-06,
|
| 920 |
+
"loss": 0.1349517822265625,
|
| 921 |
+
"step": 565,
|
| 922 |
+
"token_acc": 0.9553399628102515
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 8.263736263736265,
|
| 926 |
+
"grad_norm": 1.194259524345398,
|
| 927 |
+
"learning_rate": 4.282321120728493e-06,
|
| 928 |
+
"loss": 0.13680732250213623,
|
| 929 |
+
"step": 570,
|
| 930 |
+
"token_acc": 0.9549881039126251
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 8.336996336996338,
|
| 934 |
+
"grad_norm": 1.249259352684021,
|
| 935 |
+
"learning_rate": 4.267766952966369e-06,
|
| 936 |
+
"loss": 0.1411210536956787,
|
| 937 |
+
"step": 575,
|
| 938 |
+
"token_acc": 0.9536771307638042
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 8.41025641025641,
|
| 942 |
+
"grad_norm": 1.1520274877548218,
|
| 943 |
+
"learning_rate": 4.253091960681222e-06,
|
| 944 |
+
"loss": 0.142136287689209,
|
| 945 |
+
"step": 580,
|
| 946 |
+
"token_acc": 0.9528990913747142
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 8.483516483516484,
|
| 950 |
+
"grad_norm": 1.1933797597885132,
|
| 951 |
+
"learning_rate": 4.238297146889281e-06,
|
| 952 |
+
"loss": 0.13622112274169923,
|
| 953 |
+
"step": 585,
|
| 954 |
+
"token_acc": 0.9551286964168304
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 8.556776556776557,
|
| 958 |
+
"grad_norm": 1.2486155033111572,
|
| 959 |
+
"learning_rate": 4.2233835227964145e-06,
|
| 960 |
+
"loss": 0.14638983011245726,
|
| 961 |
+
"step": 590,
|
| 962 |
+
"token_acc": 0.9520263810341755
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 8.63003663003663,
|
| 966 |
+
"grad_norm": 1.196968913078308,
|
| 967 |
+
"learning_rate": 4.208352107729021e-06,
|
| 968 |
+
"loss": 0.14007785320281982,
|
| 969 |
+
"step": 595,
|
| 970 |
+
"token_acc": 0.9536529477016753
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 8.703296703296703,
|
| 974 |
+
"grad_norm": 1.1765471696853638,
|
| 975 |
+
"learning_rate": 4.1932039290643534e-06,
|
| 976 |
+
"loss": 0.14584920406341553,
|
| 977 |
+
"step": 600,
|
| 978 |
+
"token_acc": 0.9516696559924226
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 8.776556776556777,
|
| 982 |
+
"grad_norm": 1.669129490852356,
|
| 983 |
+
"learning_rate": 4.177940022160299e-06,
|
| 984 |
+
"loss": 0.0768501341342926,
|
| 985 |
+
"step": 605,
|
| 986 |
+
"token_acc": 0.975236159418236
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 8.84981684981685,
|
| 990 |
+
"grad_norm": 1.268017292022705,
|
| 991 |
+
"learning_rate": 4.162561430284621e-06,
|
| 992 |
+
"loss": 0.07224024534225464,
|
| 993 |
+
"step": 610,
|
| 994 |
+
"token_acc": 0.9766335806416627
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 8.923076923076923,
|
| 998 |
+
"grad_norm": 1.5425317287445068,
|
| 999 |
+
"learning_rate": 4.147069204543645e-06,
|
| 1000 |
+
"loss": 0.06993744373321534,
|
| 1001 |
+
"step": 615,
|
| 1002 |
+
"token_acc": 0.9771406165363689
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 8.996336996336996,
|
| 1006 |
+
"grad_norm": 1.223515510559082,
|
| 1007 |
+
"learning_rate": 4.1314644038104215e-06,
|
| 1008 |
+
"loss": 0.05478466749191284,
|
| 1009 |
+
"step": 620,
|
| 1010 |
+
"token_acc": 0.9822947776253128
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 9.058608058608058,
|
| 1014 |
+
"grad_norm": 1.2811061143875122,
|
| 1015 |
+
"learning_rate": 4.115748094652352e-06,
|
| 1016 |
+
"loss": 0.0926189124584198,
|
| 1017 |
+
"step": 625,
|
| 1018 |
+
"token_acc": 0.9643604931640383
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 9.131868131868131,
|
| 1022 |
+
"grad_norm": 1.280188798904419,
|
| 1023 |
+
"learning_rate": 4.099921351258292e-06,
|
| 1024 |
+
"loss": 0.10901727676391601,
|
| 1025 |
+
"step": 630,
|
| 1026 |
+
"token_acc": 0.9643512039762583
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 9.205128205128204,
|
| 1030 |
+
"grad_norm": 1.1829010248184204,
|
| 1031 |
+
"learning_rate": 4.083985255365127e-06,
|
| 1032 |
+
"loss": 0.10478531122207642,
|
| 1033 |
+
"step": 635,
|
| 1034 |
+
"token_acc": 0.9658587354642435
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 9.27838827838828,
|
| 1038 |
+
"grad_norm": 1.574622392654419,
|
| 1039 |
+
"learning_rate": 4.067940896183843e-06,
|
| 1040 |
+
"loss": 0.10349674224853515,
|
| 1041 |
+
"step": 640,
|
| 1042 |
+
"token_acc": 0.9662280454057717
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 9.351648351648352,
|
| 1046 |
+
"grad_norm": 1.148873209953308,
|
| 1047 |
+
"learning_rate": 4.051789370325079e-06,
|
| 1048 |
+
"loss": 0.10512499809265137,
|
| 1049 |
+
"step": 645,
|
| 1050 |
+
"token_acc": 0.9659510878298037
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 9.424908424908425,
|
| 1054 |
+
"grad_norm": 1.2910243272781372,
|
| 1055 |
+
"learning_rate": 4.0355317817241705e-06,
|
| 1056 |
+
"loss": 0.10366677045822144,
|
| 1057 |
+
"step": 650,
|
| 1058 |
+
"token_acc": 0.9657937126316597
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 9.498168498168498,
|
| 1062 |
+
"grad_norm": 1.2024842500686646,
|
| 1063 |
+
"learning_rate": 4.019169241565704e-06,
|
| 1064 |
+
"loss": 0.10600372552871704,
|
| 1065 |
+
"step": 655,
|
| 1066 |
+
"token_acc": 0.9653143400578701
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 9.571428571428571,
|
| 1070 |
+
"grad_norm": 1.1264986991882324,
|
| 1071 |
+
"learning_rate": 4.002702868207563e-06,
|
| 1072 |
+
"loss": 0.10574545860290527,
|
| 1073 |
+
"step": 660,
|
| 1074 |
+
"token_acc": 0.9652327184375745
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 9.644688644688644,
|
| 1078 |
+
"grad_norm": 1.1488364934921265,
|
| 1079 |
+
"learning_rate": 3.986133787104496e-06,
|
| 1080 |
+
"loss": 0.1057441234588623,
|
| 1081 |
+
"step": 665,
|
| 1082 |
+
"token_acc": 0.9653137066220019
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 9.717948717948717,
|
| 1086 |
+
"grad_norm": 1.2188081741333008,
|
| 1087 |
+
"learning_rate": 3.969463130731183e-06,
|
| 1088 |
+
"loss": 0.10680370330810547,
|
| 1089 |
+
"step": 670,
|
| 1090 |
+
"token_acc": 0.9654117675356486
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 9.791208791208792,
|
| 1094 |
+
"grad_norm": 1.225588083267212,
|
| 1095 |
+
"learning_rate": 3.9526920385048465e-06,
|
| 1096 |
+
"loss": 0.1085613489151001,
|
| 1097 |
+
"step": 675,
|
| 1098 |
+
"token_acc": 0.9642317155724182
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 9.864468864468865,
|
| 1102 |
+
"grad_norm": 1.1125155687332153,
|
| 1103 |
+
"learning_rate": 3.935821656707359e-06,
|
| 1104 |
+
"loss": 0.10585691928863525,
|
| 1105 |
+
"step": 680,
|
| 1106 |
+
"token_acc": 0.965347073112761
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 9.937728937728938,
|
| 1110 |
+
"grad_norm": 1.1567717790603638,
|
| 1111 |
+
"learning_rate": 3.9188531384069095e-06,
|
| 1112 |
+
"loss": 0.10892641544342041,
|
| 1113 |
+
"step": 685,
|
| 1114 |
+
"token_acc": 0.9641639469059516
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 10.0,
|
| 1118 |
+
"grad_norm": 3.9209771156311035,
|
| 1119 |
+
"learning_rate": 3.901787643379183e-06,
|
| 1120 |
+
"loss": 0.1068874478340149,
|
| 1121 |
+
"step": 690,
|
| 1122 |
+
"token_acc": 0.9642589277678331
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 10.073260073260073,
|
| 1126 |
+
"grad_norm": 1.167458415031433,
|
| 1127 |
+
"learning_rate": 3.884626338028093e-06,
|
| 1128 |
+
"loss": 0.06645301580429078,
|
| 1129 |
+
"step": 695,
|
| 1130 |
+
"token_acc": 0.9785720640209629
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 10.146520146520146,
|
| 1134 |
+
"grad_norm": 1.0500937700271606,
|
| 1135 |
+
"learning_rate": 3.8673703953060685e-06,
|
| 1136 |
+
"loss": 0.06310778856277466,
|
| 1137 |
+
"step": 700,
|
| 1138 |
+
"token_acc": 0.9799111384227273
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 10.219780219780219,
|
| 1142 |
+
"grad_norm": 1.1901004314422607,
|
| 1143 |
+
"learning_rate": 3.850020994633869e-06,
|
| 1144 |
+
"loss": 0.09779422283172608,
|
| 1145 |
+
"step": 705,
|
| 1146 |
+
"token_acc": 0.968069333783344
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 10.293040293040294,
|
| 1150 |
+
"grad_norm": 1.040907859802246,
|
| 1151 |
+
"learning_rate": 3.832579321819985e-06,
|
| 1152 |
+
"loss": 0.08124626278877259,
|
| 1153 |
+
"step": 710,
|
| 1154 |
+
"token_acc": 0.9736844221105527
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 10.366300366300367,
|
| 1158 |
+
"grad_norm": 1.8486651182174683,
|
| 1159 |
+
"learning_rate": 3.815046568979585e-06,
|
| 1160 |
+
"loss": 0.06958885192871093,
|
| 1161 |
+
"step": 715,
|
| 1162 |
+
"token_acc": 0.9773133724458831
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 10.43956043956044,
|
| 1166 |
+
"grad_norm": 1.110428810119629,
|
| 1167 |
+
"learning_rate": 3.797423934453038e-06,
|
| 1168 |
+
"loss": 0.061410021781921384,
|
| 1169 |
+
"step": 720,
|
| 1170 |
+
"token_acc": 0.9802959042909349
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 10.512820512820513,
|
| 1174 |
+
"grad_norm": 1.1162679195404053,
|
| 1175 |
+
"learning_rate": 3.779712622724003e-06,
|
| 1176 |
+
"loss": 0.05629628896713257,
|
| 1177 |
+
"step": 725,
|
| 1178 |
+
"token_acc": 0.9817005456037854
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 10.586080586080586,
|
| 1182 |
+
"grad_norm": 1.1255736351013184,
|
| 1183 |
+
"learning_rate": 3.76191384433711e-06,
|
| 1184 |
+
"loss": 0.05823663473129272,
|
| 1185 |
+
"step": 730,
|
| 1186 |
+
"token_acc": 0.9810973319818929
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 10.659340659340659,
|
| 1190 |
+
"grad_norm": 1.2661657333374023,
|
| 1191 |
+
"learning_rate": 3.744028815815219e-06,
|
| 1192 |
+
"loss": 0.0595287024974823,
|
| 1193 |
+
"step": 735,
|
| 1194 |
+
"token_acc": 0.9805077554762872
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 10.732600732600732,
|
| 1198 |
+
"grad_norm": 0.9871035218238831,
|
| 1199 |
+
"learning_rate": 3.726058759576271e-06,
|
| 1200 |
+
"loss": 0.05217963457107544,
|
| 1201 |
+
"step": 740,
|
| 1202 |
+
"token_acc": 0.9830647045082924
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 10.805860805860807,
|
| 1206 |
+
"grad_norm": 1.176674723625183,
|
| 1207 |
+
"learning_rate": 3.7080049038497405e-06,
|
| 1208 |
+
"loss": 0.0421576201915741,
|
| 1209 |
+
"step": 745,
|
| 1210 |
+
"token_acc": 0.9865278628291488
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 10.87912087912088,
|
| 1214 |
+
"grad_norm": 1.1151669025421143,
|
| 1215 |
+
"learning_rate": 3.6898684825926845e-06,
|
| 1216 |
+
"loss": 0.04041574597358703,
|
| 1217 |
+
"step": 750,
|
| 1218 |
+
"token_acc": 0.9872045372718925
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 10.952380952380953,
|
| 1222 |
+
"grad_norm": 1.1017118692398071,
|
| 1223 |
+
"learning_rate": 3.6716507354054044e-06,
|
| 1224 |
+
"loss": 0.03693513870239258,
|
| 1225 |
+
"step": 755,
|
| 1226 |
+
"token_acc": 0.9883869408471129
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 11.014652014652015,
|
| 1230 |
+
"grad_norm": 1.4862247705459595,
|
| 1231 |
+
"learning_rate": 3.65335290744672e-06,
|
| 1232 |
+
"loss": 0.03411419987678528,
|
| 1233 |
+
"step": 760,
|
| 1234 |
+
"token_acc": 0.9875135964716225
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 11.087912087912088,
|
| 1238 |
+
"grad_norm": 1.46209716796875,
|
| 1239 |
+
"learning_rate": 3.634976249348867e-06,
|
| 1240 |
+
"loss": 0.05428239703178406,
|
| 1241 |
+
"step": 765,
|
| 1242 |
+
"token_acc": 0.982443522286019
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 11.16117216117216,
|
| 1246 |
+
"grad_norm": 1.0127854347229004,
|
| 1247 |
+
"learning_rate": 3.616522017132017e-06,
|
| 1248 |
+
"loss": 0.04964028000831604,
|
| 1249 |
+
"step": 770,
|
| 1250 |
+
"token_acc": 0.9845350844640927
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 11.234432234432234,
|
| 1254 |
+
"grad_norm": 1.1573317050933838,
|
| 1255 |
+
"learning_rate": 3.5979914721184263e-06,
|
| 1256 |
+
"loss": 0.05009516477584839,
|
| 1257 |
+
"step": 775,
|
| 1258 |
+
"token_acc": 0.984208285126623
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 11.307692307692308,
|
| 1262 |
+
"grad_norm": 1.0386983156204224,
|
| 1263 |
+
"learning_rate": 3.579385880846232e-06,
|
| 1264 |
+
"loss": 0.05456154942512512,
|
| 1265 |
+
"step": 780,
|
| 1266 |
+
"token_acc": 0.9827983505821475
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 11.380952380952381,
|
| 1270 |
+
"grad_norm": 1.2116971015930176,
|
| 1271 |
+
"learning_rate": 3.5607065149828845e-06,
|
| 1272 |
+
"loss": 0.05355601310729981,
|
| 1273 |
+
"step": 785,
|
| 1274 |
+
"token_acc": 0.982528596011742
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 11.454212454212454,
|
| 1278 |
+
"grad_norm": 1.163842797279358,
|
| 1279 |
+
"learning_rate": 3.5419546512382264e-06,
|
| 1280 |
+
"loss": 0.05224733352661133,
|
| 1281 |
+
"step": 790,
|
| 1282 |
+
"token_acc": 0.9832931720445247
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 11.527472527472527,
|
| 1286 |
+
"grad_norm": 1.1067230701446533,
|
| 1287 |
+
"learning_rate": 3.523131571277235e-06,
|
| 1288 |
+
"loss": 0.05252179503440857,
|
| 1289 |
+
"step": 795,
|
| 1290 |
+
"token_acc": 0.9834636187530457
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 11.6007326007326,
|
| 1294 |
+
"grad_norm": 1.193214774131775,
|
| 1295 |
+
"learning_rate": 3.5042385616324243e-06,
|
| 1296 |
+
"loss": 0.051275789737701416,
|
| 1297 |
+
"step": 800,
|
| 1298 |
+
"token_acc": 0.9835009377223048
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 11.673992673992673,
|
| 1302 |
+
"grad_norm": 1.21199369430542,
|
| 1303 |
+
"learning_rate": 3.4852769136159047e-06,
|
| 1304 |
+
"loss": 0.03126579523086548,
|
| 1305 |
+
"step": 805,
|
| 1306 |
+
"token_acc": 0.990717682593747
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 11.747252747252748,
|
| 1310 |
+
"grad_norm": 0.9968395829200745,
|
| 1311 |
+
"learning_rate": 3.466247923231131e-06,
|
| 1312 |
+
"loss": 0.026918619871139526,
|
| 1313 |
+
"step": 810,
|
| 1314 |
+
"token_acc": 0.9920119860593407
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 11.820512820512821,
|
| 1318 |
+
"grad_norm": 1.2335270643234253,
|
| 1319 |
+
"learning_rate": 3.4471528910843193e-06,
|
| 1320 |
+
"loss": 0.025720825791358946,
|
| 1321 |
+
"step": 815,
|
| 1322 |
+
"token_acc": 0.9922731754431857
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 11.893772893772894,
|
| 1326 |
+
"grad_norm": 1.3252135515213013,
|
| 1327 |
+
"learning_rate": 3.427993122295552e-06,
|
| 1328 |
+
"loss": 0.02836277484893799,
|
| 1329 |
+
"step": 820,
|
| 1330 |
+
"token_acc": 0.991218019053943
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 11.967032967032967,
|
| 1334 |
+
"grad_norm": 1.0112091302871704,
|
| 1335 |
+
"learning_rate": 3.4087699264095746e-06,
|
| 1336 |
+
"loss": 0.02677382230758667,
|
| 1337 |
+
"step": 825,
|
| 1338 |
+
"token_acc": 0.9920123284465996
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 12.02930402930403,
|
| 1342 |
+
"grad_norm": 1.9008547067642212,
|
| 1343 |
+
"learning_rate": 3.3894846173062917e-06,
|
| 1344 |
+
"loss": 0.035012418031692506,
|
| 1345 |
+
"step": 830,
|
| 1346 |
+
"token_acc": 0.9882785008376039
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 12.102564102564102,
|
| 1350 |
+
"grad_norm": 1.3940945863723755,
|
| 1351 |
+
"learning_rate": 3.3701385131109617e-06,
|
| 1352 |
+
"loss": 0.04694617986679077,
|
| 1353 |
+
"step": 835,
|
| 1354 |
+
"token_acc": 0.9850311732782013
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 12.175824175824175,
|
| 1358 |
+
"grad_norm": 1.114364743232727,
|
| 1359 |
+
"learning_rate": 3.350732936104108e-06,
|
| 1360 |
+
"loss": 0.04108467400074005,
|
| 1361 |
+
"step": 840,
|
| 1362 |
+
"token_acc": 0.9870112080704553
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 12.249084249084248,
|
| 1366 |
+
"grad_norm": 1.0285835266113281,
|
| 1367 |
+
"learning_rate": 3.3312692126311424e-06,
|
| 1368 |
+
"loss": 0.04414758384227753,
|
| 1369 |
+
"step": 845,
|
| 1370 |
+
"token_acc": 0.9861624663488048
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 12.322344322344323,
|
| 1374 |
+
"grad_norm": 0.9784775376319885,
|
| 1375 |
+
"learning_rate": 3.3117486730117092e-06,
|
| 1376 |
+
"loss": 0.04183251857757568,
|
| 1377 |
+
"step": 850,
|
| 1378 |
+
"token_acc": 0.9869409616751761
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 12.395604395604396,
|
| 1382 |
+
"grad_norm": 0.9582324028015137,
|
| 1383 |
+
"learning_rate": 3.292172651448761e-06,
|
| 1384 |
+
"loss": 0.04124728441238403,
|
| 1385 |
+
"step": 855,
|
| 1386 |
+
"token_acc": 0.9869587855646551
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 12.468864468864469,
|
| 1390 |
+
"grad_norm": 1.0609194040298462,
|
| 1391 |
+
"learning_rate": 3.272542485937369e-06,
|
| 1392 |
+
"loss": 0.040315544605255126,
|
| 1393 |
+
"step": 860,
|
| 1394 |
+
"token_acc": 0.9875006531009232
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 12.542124542124542,
|
| 1398 |
+
"grad_norm": 0.9905833005905151,
|
| 1399 |
+
"learning_rate": 3.252859518173269e-06,
|
| 1400 |
+
"loss": 0.04133622050285339,
|
| 1401 |
+
"step": 865,
|
| 1402 |
+
"token_acc": 0.9868536160581716
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 12.615384615384615,
|
| 1406 |
+
"grad_norm": 1.1323637962341309,
|
| 1407 |
+
"learning_rate": 3.2331250934611623e-06,
|
| 1408 |
+
"loss": 0.04258134961128235,
|
| 1409 |
+
"step": 870,
|
| 1410 |
+
"token_acc": 0.9864731298184181
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 12.688644688644688,
|
| 1414 |
+
"grad_norm": 1.0161943435668945,
|
| 1415 |
+
"learning_rate": 3.2133405606227636e-06,
|
| 1416 |
+
"loss": 0.0409242033958435,
|
| 1417 |
+
"step": 875,
|
| 1418 |
+
"token_acc": 0.9871906594692293
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 12.761904761904763,
|
| 1422 |
+
"grad_norm": 1.0795347690582275,
|
| 1423 |
+
"learning_rate": 3.193507271904612e-06,
|
| 1424 |
+
"loss": 0.04416788816452026,
|
| 1425 |
+
"step": 880,
|
| 1426 |
+
"token_acc": 0.9860561712856525
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 12.835164835164836,
|
| 1430 |
+
"grad_norm": 1.0203306674957275,
|
| 1431 |
+
"learning_rate": 3.173626582885645e-06,
|
| 1432 |
+
"loss": 0.04076763093471527,
|
| 1433 |
+
"step": 885,
|
| 1434 |
+
"token_acc": 0.9870355935557611
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 12.908424908424909,
|
| 1438 |
+
"grad_norm": 1.0302674770355225,
|
| 1439 |
+
"learning_rate": 3.15369985238455e-06,
|
| 1440 |
+
"loss": 0.04232300519943237,
|
| 1441 |
+
"step": 890,
|
| 1442 |
+
"token_acc": 0.9867237962788042
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 12.981684981684982,
|
| 1446 |
+
"grad_norm": 1.103152871131897,
|
| 1447 |
+
"learning_rate": 3.133728442366885e-06,
|
| 1448 |
+
"loss": 0.044867289066314694,
|
| 1449 |
+
"step": 895,
|
| 1450 |
+
"token_acc": 0.9857875446033484
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 13.043956043956044,
|
| 1454 |
+
"grad_norm": 0.8937996029853821,
|
| 1455 |
+
"learning_rate": 3.1137137178519983e-06,
|
| 1456 |
+
"loss": 0.03286854326725006,
|
| 1457 |
+
"step": 900,
|
| 1458 |
+
"token_acc": 0.9904325136395983
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 13.117216117216117,
|
| 1462 |
+
"grad_norm": 1.1743171215057373,
|
| 1463 |
+
"learning_rate": 3.093657046819722e-06,
|
| 1464 |
+
"loss": 0.04345293045043945,
|
| 1465 |
+
"step": 905,
|
| 1466 |
+
"token_acc": 0.9861550840844119
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 13.19047619047619,
|
| 1470 |
+
"grad_norm": 0.9643164277076721,
|
| 1471 |
+
"learning_rate": 3.073559800116879e-06,
|
| 1472 |
+
"loss": 0.038184982538223264,
|
| 1473 |
+
"step": 910,
|
| 1474 |
+
"token_acc": 0.9882567849686847
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 13.263736263736265,
|
| 1478 |
+
"grad_norm": 1.0102027654647827,
|
| 1479 |
+
"learning_rate": 3.0534233513635863e-06,
|
| 1480 |
+
"loss": 0.03169253766536713,
|
| 1481 |
+
"step": 915,
|
| 1482 |
+
"token_acc": 0.9902852526356829
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 13.336996336996338,
|
| 1486 |
+
"grad_norm": 0.9551713466644287,
|
| 1487 |
+
"learning_rate": 3.0332490768593676e-06,
|
| 1488 |
+
"loss": 0.028652000427246093,
|
| 1489 |
+
"step": 920,
|
| 1490 |
+
"token_acc": 0.9912297025056837
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 13.41025641025641,
|
| 1494 |
+
"grad_norm": 0.8975064158439636,
|
| 1495 |
+
"learning_rate": 3.013038355489086e-06,
|
| 1496 |
+
"loss": 0.027258700132369994,
|
| 1497 |
+
"step": 925,
|
| 1498 |
+
"token_acc": 0.9917707954223786
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 13.483516483516484,
|
| 1502 |
+
"grad_norm": 0.9920509457588196,
|
| 1503 |
+
"learning_rate": 2.9927925686287006e-06,
|
| 1504 |
+
"loss": 0.026785725355148317,
|
| 1505 |
+
"step": 930,
|
| 1506 |
+
"token_acc": 0.9918329906944372
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 13.556776556776557,
|
| 1510 |
+
"grad_norm": 0.9444341063499451,
|
| 1511 |
+
"learning_rate": 2.972513100050851e-06,
|
| 1512 |
+
"loss": 0.027184110879898072,
|
| 1513 |
+
"step": 935,
|
| 1514 |
+
"token_acc": 0.9917801813278293
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 13.63003663003663,
|
| 1518 |
+
"grad_norm": 0.9316787719726562,
|
| 1519 |
+
"learning_rate": 2.9522013358302754e-06,
|
| 1520 |
+
"loss": 0.023839214444160463,
|
| 1521 |
+
"step": 940,
|
| 1522 |
+
"token_acc": 0.9928025444412724
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 13.703296703296703,
|
| 1526 |
+
"grad_norm": 0.9105445742607117,
|
| 1527 |
+
"learning_rate": 2.9318586642490766e-06,
|
| 1528 |
+
"loss": 0.0220517098903656,
|
| 1529 |
+
"step": 945,
|
| 1530 |
+
"token_acc": 0.9935465701066184
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 13.776556776556777,
|
| 1534 |
+
"grad_norm": 1.0082361698150635,
|
| 1535 |
+
"learning_rate": 2.911486475701835e-06,
|
| 1536 |
+
"loss": 0.018666207790374756,
|
| 1537 |
+
"step": 950,
|
| 1538 |
+
"token_acc": 0.9945095212837031
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 13.84981684981685,
|
| 1542 |
+
"grad_norm": 1.0385477542877197,
|
| 1543 |
+
"learning_rate": 2.8910861626005774e-06,
|
| 1544 |
+
"loss": 0.02035895437002182,
|
| 1545 |
+
"step": 955,
|
| 1546 |
+
"token_acc": 0.9940067564582019
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 13.923076923076923,
|
| 1550 |
+
"grad_norm": 1.0480846166610718,
|
| 1551 |
+
"learning_rate": 2.870659119279605e-06,
|
| 1552 |
+
"loss": 0.020590776205062868,
|
| 1553 |
+
"step": 960,
|
| 1554 |
+
"token_acc": 0.9938516141028854
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 13.996336996336996,
|
| 1558 |
+
"grad_norm": 0.9408004879951477,
|
| 1559 |
+
"learning_rate": 2.850206741900195e-06,
|
| 1560 |
+
"loss": 0.018829339742660524,
|
| 1561 |
+
"step": 965,
|
| 1562 |
+
"token_acc": 0.9943215755912503
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 14.058608058608058,
|
| 1566 |
+
"grad_norm": 0.9699022769927979,
|
| 1567 |
+
"learning_rate": 2.829730428355173e-06,
|
| 1568 |
+
"loss": 0.02413397878408432,
|
| 1569 |
+
"step": 970,
|
| 1570 |
+
"token_acc": 0.9913756553055432
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 14.131868131868131,
|
| 1574 |
+
"grad_norm": 0.9755733609199524,
|
| 1575 |
+
"learning_rate": 2.80923157817337e-06,
|
| 1576 |
+
"loss": 0.024001887440681456,
|
| 1577 |
+
"step": 975,
|
| 1578 |
+
"token_acc": 0.9928639006363864
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 14.205128205128204,
|
| 1582 |
+
"grad_norm": 0.9670774936676025,
|
| 1583 |
+
"learning_rate": 2.788711592423966e-06,
|
| 1584 |
+
"loss": 0.022572694718837737,
|
| 1585 |
+
"step": 980,
|
| 1586 |
+
"token_acc": 0.993273551621066
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 14.27838827838828,
|
| 1590 |
+
"grad_norm": 0.8622258901596069,
|
| 1591 |
+
"learning_rate": 2.76817187362073e-06,
|
| 1592 |
+
"loss": 0.02251937985420227,
|
| 1593 |
+
"step": 985,
|
| 1594 |
+
"token_acc": 0.993419069021383
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 14.351648351648352,
|
| 1598 |
+
"grad_norm": 0.9087955951690674,
|
| 1599 |
+
"learning_rate": 2.7476138256261575e-06,
|
| 1600 |
+
"loss": 0.022449295222759246,
|
| 1601 |
+
"step": 990,
|
| 1602 |
+
"token_acc": 0.9932415077377512
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 14.424908424908425,
|
| 1606 |
+
"grad_norm": 0.8483551144599915,
|
| 1607 |
+
"learning_rate": 2.7270388535555207e-06,
|
| 1608 |
+
"loss": 0.022757737338542937,
|
| 1609 |
+
"step": 995,
|
| 1610 |
+
"token_acc": 0.9932496614725439
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 14.498168498168498,
|
| 1614 |
+
"grad_norm": 0.8271490931510925,
|
| 1615 |
+
"learning_rate": 2.7064483636808314e-06,
|
| 1616 |
+
"loss": 0.02226755619049072,
|
| 1617 |
+
"step": 1000,
|
| 1618 |
+
"token_acc": 0.9932772431047354
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 14.571428571428571,
|
| 1622 |
+
"grad_norm": 0.8335503339767456,
|
| 1623 |
+
"learning_rate": 2.6858437633347197e-06,
|
| 1624 |
+
"loss": 0.015954886376857758,
|
| 1625 |
+
"step": 1005,
|
| 1626 |
+
"token_acc": 0.9955423476968797
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 14.644688644688644,
|
| 1630 |
+
"grad_norm": 0.7704828381538391,
|
| 1631 |
+
"learning_rate": 2.6652264608142487e-06,
|
| 1632 |
+
"loss": 0.014800366759300233,
|
| 1633 |
+
"step": 1010,
|
| 1634 |
+
"token_acc": 0.9956421152409488
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 14.717948717948717,
|
| 1638 |
+
"grad_norm": 0.872599720954895,
|
| 1639 |
+
"learning_rate": 2.6445978652846605e-06,
|
| 1640 |
+
"loss": 0.015232875943183899,
|
| 1641 |
+
"step": 1015,
|
| 1642 |
+
"token_acc": 0.9958409923697896
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 14.791208791208792,
|
| 1646 |
+
"grad_norm": 0.8955268859863281,
|
| 1647 |
+
"learning_rate": 2.623959386683056e-06,
|
| 1648 |
+
"loss": 0.015665414929389953,
|
| 1649 |
+
"step": 1020,
|
| 1650 |
+
"token_acc": 0.9953696589319077
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 14.864468864468865,
|
| 1654 |
+
"grad_norm": 1.0283321142196655,
|
| 1655 |
+
"learning_rate": 2.603312435622033e-06,
|
| 1656 |
+
"loss": 0.01725401431322098,
|
| 1657 |
+
"step": 1025,
|
| 1658 |
+
"token_acc": 0.9950444313153357
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 14.937728937728938,
|
| 1662 |
+
"grad_norm": 1.1370958089828491,
|
| 1663 |
+
"learning_rate": 2.5826584232932707e-06,
|
| 1664 |
+
"loss": 0.018383359909057616,
|
| 1665 |
+
"step": 1030,
|
| 1666 |
+
"token_acc": 0.9945760822593869
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 15.0,
|
| 1670 |
+
"grad_norm": 2.657243251800537,
|
| 1671 |
+
"learning_rate": 2.5619987613710757e-06,
|
| 1672 |
+
"loss": 0.014658606052398682,
|
| 1673 |
+
"step": 1035,
|
| 1674 |
+
"token_acc": 0.9951515996236765
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 15.073260073260073,
|
| 1678 |
+
"grad_norm": 1.0341904163360596,
|
| 1679 |
+
"learning_rate": 2.5413348619158966e-06,
|
| 1680 |
+
"loss": 0.02338849604129791,
|
| 1681 |
+
"step": 1040,
|
| 1682 |
+
"token_acc": 0.9926503867058112
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 15.146520146520146,
|
| 1686 |
+
"grad_norm": 0.9077109098434448,
|
| 1687 |
+
"learning_rate": 2.5206681372778126e-06,
|
| 1688 |
+
"loss": 0.022106045484542848,
|
| 1689 |
+
"step": 1045,
|
| 1690 |
+
"token_acc": 0.9934412334517275
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 15.219780219780219,
|
| 1694 |
+
"grad_norm": 0.9011284112930298,
|
| 1695 |
+
"learning_rate": 2.5e-06,
|
| 1696 |
+
"loss": 0.021497051417827606,
|
| 1697 |
+
"step": 1050,
|
| 1698 |
+
"token_acc": 0.99361983135223
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 15.293040293040294,
|
| 1702 |
+
"grad_norm": 0.8781850934028625,
|
| 1703 |
+
"learning_rate": 2.479331862722188e-06,
|
| 1704 |
+
"loss": 0.020798739790916444,
|
| 1705 |
+
"step": 1055,
|
| 1706 |
+
"token_acc": 0.9940434406839493
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 15.366300366300367,
|
| 1710 |
+
"grad_norm": 0.7928971648216248,
|
| 1711 |
+
"learning_rate": 2.458665138084104e-06,
|
| 1712 |
+
"loss": 0.021296033263206483,
|
| 1713 |
+
"step": 1060,
|
| 1714 |
+
"token_acc": 0.9936675651968422
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 15.43956043956044,
|
| 1718 |
+
"grad_norm": 0.9471123814582825,
|
| 1719 |
+
"learning_rate": 2.438001238628925e-06,
|
| 1720 |
+
"loss": 0.021688148379325867,
|
| 1721 |
+
"step": 1065,
|
| 1722 |
+
"token_acc": 0.9934400419837313
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 15.512820512820513,
|
| 1726 |
+
"grad_norm": 0.8935464024543762,
|
| 1727 |
+
"learning_rate": 2.4173415767067297e-06,
|
| 1728 |
+
"loss": 0.02238973379135132,
|
| 1729 |
+
"step": 1070,
|
| 1730 |
+
"token_acc": 0.9933113076150681
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 15.586080586080586,
|
| 1734 |
+
"grad_norm": 0.9456347823143005,
|
| 1735 |
+
"learning_rate": 2.396687564377967e-06,
|
| 1736 |
+
"loss": 0.020350757241249084,
|
| 1737 |
+
"step": 1075,
|
| 1738 |
+
"token_acc": 0.9939768789870794
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 15.659340659340659,
|
| 1742 |
+
"grad_norm": 0.7795377969741821,
|
| 1743 |
+
"learning_rate": 2.376040613316944e-06,
|
| 1744 |
+
"loss": 0.021586185693740843,
|
| 1745 |
+
"step": 1080,
|
| 1746 |
+
"token_acc": 0.9934239987735556
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 15.732600732600732,
|
| 1750 |
+
"grad_norm": 0.8604532480239868,
|
| 1751 |
+
"learning_rate": 2.3554021347153403e-06,
|
| 1752 |
+
"loss": 0.021918649971485137,
|
| 1753 |
+
"step": 1085,
|
| 1754 |
+
"token_acc": 0.9933028540206981
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 15.805860805860807,
|
| 1758 |
+
"grad_norm": 0.8883838653564453,
|
| 1759 |
+
"learning_rate": 2.3347735391857517e-06,
|
| 1760 |
+
"loss": 0.02288481891155243,
|
| 1761 |
+
"step": 1090,
|
| 1762 |
+
"token_acc": 0.9929645568396113
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 15.87912087912088,
|
| 1766 |
+
"grad_norm": 0.8979478478431702,
|
| 1767 |
+
"learning_rate": 2.3141562366652816e-06,
|
| 1768 |
+
"loss": 0.02122913599014282,
|
| 1769 |
+
"step": 1095,
|
| 1770 |
+
"token_acc": 0.993564648039141
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 15.952380952380953,
|
| 1774 |
+
"grad_norm": 0.8900680541992188,
|
| 1775 |
+
"learning_rate": 2.2935516363191695e-06,
|
| 1776 |
+
"loss": 0.022361186146736146,
|
| 1777 |
+
"step": 1100,
|
| 1778 |
+
"token_acc": 0.9933146256230841
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 16.014652014652015,
|
| 1782 |
+
"grad_norm": 0.9495336413383484,
|
| 1783 |
+
"learning_rate": 2.2729611464444797e-06,
|
| 1784 |
+
"loss": 0.009362544119358062,
|
| 1785 |
+
"step": 1105,
|
| 1786 |
+
"token_acc": 0.9972890292735096
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 16.087912087912088,
|
| 1790 |
+
"grad_norm": 0.8308697938919067,
|
| 1791 |
+
"learning_rate": 2.2523861743738433e-06,
|
| 1792 |
+
"loss": 0.016145363450050354,
|
| 1793 |
+
"step": 1110,
|
| 1794 |
+
"token_acc": 0.9954890077435329
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 16.16117216117216,
|
| 1798 |
+
"grad_norm": 0.7013668417930603,
|
| 1799 |
+
"learning_rate": 2.2318281263792714e-06,
|
| 1800 |
+
"loss": 0.014486923813819885,
|
| 1801 |
+
"step": 1115,
|
| 1802 |
+
"token_acc": 0.9959613325980472
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 16.234432234432234,
|
| 1806 |
+
"grad_norm": 0.7565985918045044,
|
| 1807 |
+
"learning_rate": 2.211288407576035e-06,
|
| 1808 |
+
"loss": 0.014733286201953888,
|
| 1809 |
+
"step": 1120,
|
| 1810 |
+
"token_acc": 0.9959043845278542
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 16.307692307692307,
|
| 1814 |
+
"grad_norm": 0.8071579337120056,
|
| 1815 |
+
"learning_rate": 2.190768421826631e-06,
|
| 1816 |
+
"loss": 0.015736982226371765,
|
| 1817 |
+
"step": 1125,
|
| 1818 |
+
"token_acc": 0.9955398804978851
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 16.38095238095238,
|
| 1822 |
+
"grad_norm": 0.6978721022605896,
|
| 1823 |
+
"learning_rate": 2.1702695716448276e-06,
|
| 1824 |
+
"loss": 0.015514859557151794,
|
| 1825 |
+
"step": 1130,
|
| 1826 |
+
"token_acc": 0.9955823470588472
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 16.454212454212453,
|
| 1830 |
+
"grad_norm": 0.7473545670509338,
|
| 1831 |
+
"learning_rate": 2.1497932580998055e-06,
|
| 1832 |
+
"loss": 0.016324551403522493,
|
| 1833 |
+
"step": 1135,
|
| 1834 |
+
"token_acc": 0.9952275500884984
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 16.52747252747253,
|
| 1838 |
+
"grad_norm": 0.7756368517875671,
|
| 1839 |
+
"learning_rate": 2.129340880720395e-06,
|
| 1840 |
+
"loss": 0.016091182827949524,
|
| 1841 |
+
"step": 1140,
|
| 1842 |
+
"token_acc": 0.995226624558916
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 16.600732600732602,
|
| 1846 |
+
"grad_norm": 0.8048065304756165,
|
| 1847 |
+
"learning_rate": 2.1089138373994226e-06,
|
| 1848 |
+
"loss": 0.017841890454292297,
|
| 1849 |
+
"step": 1145,
|
| 1850 |
+
"token_acc": 0.9948390426424102
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 16.673992673992675,
|
| 1854 |
+
"grad_norm": 0.8308281302452087,
|
| 1855 |
+
"learning_rate": 2.088513524298165e-06,
|
| 1856 |
+
"loss": 0.017377564311027528,
|
| 1857 |
+
"step": 1150,
|
| 1858 |
+
"token_acc": 0.9950155386043508
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 16.747252747252748,
|
| 1862 |
+
"grad_norm": 0.7720639705657959,
|
| 1863 |
+
"learning_rate": 2.068141335750925e-06,
|
| 1864 |
+
"loss": 0.016499611735343932,
|
| 1865 |
+
"step": 1155,
|
| 1866 |
+
"token_acc": 0.9953219382608381
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 16.82051282051282,
|
| 1870 |
+
"grad_norm": 0.8895593881607056,
|
| 1871 |
+
"learning_rate": 2.0477986641697263e-06,
|
| 1872 |
+
"loss": 0.017675170302391054,
|
| 1873 |
+
"step": 1160,
|
| 1874 |
+
"token_acc": 0.9948620353903967
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 16.893772893772894,
|
| 1878 |
+
"grad_norm": 0.8066741228103638,
|
| 1879 |
+
"learning_rate": 2.0274868999491496e-06,
|
| 1880 |
+
"loss": 0.017519712448120117,
|
| 1881 |
+
"step": 1165,
|
| 1882 |
+
"token_acc": 0.9946325234398629
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 16.967032967032967,
|
| 1886 |
+
"grad_norm": 0.9140301942825317,
|
| 1887 |
+
"learning_rate": 2.0072074313713e-06,
|
| 1888 |
+
"loss": 0.018758855760097504,
|
| 1889 |
+
"step": 1170,
|
| 1890 |
+
"token_acc": 0.9942774529228886
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 17.02930402930403,
|
| 1894 |
+
"grad_norm": 0.7352685332298279,
|
| 1895 |
+
"learning_rate": 1.9869616445109146e-06,
|
| 1896 |
+
"loss": 0.014990274608135224,
|
| 1897 |
+
"step": 1175,
|
| 1898 |
+
"token_acc": 0.995614100443557
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 17.102564102564102,
|
| 1902 |
+
"grad_norm": 0.7146985530853271,
|
| 1903 |
+
"learning_rate": 1.9667509231406332e-06,
|
| 1904 |
+
"loss": 0.012896986305713653,
|
| 1905 |
+
"step": 1180,
|
| 1906 |
+
"token_acc": 0.9964796257838175
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 17.175824175824175,
|
| 1910 |
+
"grad_norm": 0.7245467305183411,
|
| 1911 |
+
"learning_rate": 1.9465766486364145e-06,
|
| 1912 |
+
"loss": 0.013568750023841858,
|
| 1913 |
+
"step": 1185,
|
| 1914 |
+
"token_acc": 0.9964279777326376
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 17.249084249084248,
|
| 1918 |
+
"grad_norm": 0.7379946112632751,
|
| 1919 |
+
"learning_rate": 1.9264401998831213e-06,
|
| 1920 |
+
"loss": 0.013740572333335876,
|
| 1921 |
+
"step": 1190,
|
| 1922 |
+
"token_acc": 0.996070118419359
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 17.32234432234432,
|
| 1926 |
+
"grad_norm": 0.7395389676094055,
|
| 1927 |
+
"learning_rate": 1.9063429531802788e-06,
|
| 1928 |
+
"loss": 0.013096806406974793,
|
| 1929 |
+
"step": 1195,
|
| 1930 |
+
"token_acc": 0.9962539114164973
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 17.395604395604394,
|
| 1934 |
+
"grad_norm": 0.6991714835166931,
|
| 1935 |
+
"learning_rate": 1.8862862821480023e-06,
|
| 1936 |
+
"loss": 0.014323487877845764,
|
| 1937 |
+
"step": 1200,
|
| 1938 |
+
"token_acc": 0.995923222016203
|
| 1939 |
+
}
|
| 1940 |
+
],
|
| 1941 |
+
"logging_steps": 5,
|
| 1942 |
+
"max_steps": 2000,
|
| 1943 |
+
"num_input_tokens_seen": 0,
|
| 1944 |
+
"num_train_epochs": 29,
|
| 1945 |
+
"save_steps": 100,
|
| 1946 |
+
"stateful_callbacks": {
|
| 1947 |
+
"TrainerControl": {
|
| 1948 |
+
"args": {
|
| 1949 |
+
"should_epoch_stop": false,
|
| 1950 |
+
"should_evaluate": false,
|
| 1951 |
+
"should_log": false,
|
| 1952 |
+
"should_save": true,
|
| 1953 |
+
"should_training_stop": false
|
| 1954 |
+
},
|
| 1955 |
+
"attributes": {}
|
| 1956 |
+
}
|
| 1957 |
+
},
|
| 1958 |
+
"total_flos": 2.815806405817139e+16,
|
| 1959 |
+
"train_batch_size": 4,
|
| 1960 |
+
"trial_name": null,
|
| 1961 |
+
"trial_params": null
|
| 1962 |
+
}
|
VA-Judger/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a853db6bdc59ad36aee90f474fadc4a3b1ff4d8a6c76ff8447e39b649536b2d
|
| 3 |
+
size 9425
|
VA-Judger/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
VA-Judger/zero_to_fp32.py
ADDED
|
@@ -0,0 +1,760 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python
|
| 2 |
+
|
| 3 |
+
# Copyright (c) Microsoft Corporation.
|
| 4 |
+
# SPDX-License-Identifier: Apache-2.0
|
| 5 |
+
|
| 6 |
+
# DeepSpeed Team
|
| 7 |
+
|
| 8 |
+
# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
|
| 9 |
+
# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
|
| 10 |
+
# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
|
| 11 |
+
# application.
|
| 12 |
+
#
|
| 13 |
+
# example:
|
| 14 |
+
# python zero_to_fp32.py . output_dir/
|
| 15 |
+
# or
|
| 16 |
+
# python zero_to_fp32.py . output_dir/ --safe_serialization
|
| 17 |
+
|
| 18 |
+
import argparse
|
| 19 |
+
import torch
|
| 20 |
+
import glob
|
| 21 |
+
import math
|
| 22 |
+
import os
|
| 23 |
+
import re
|
| 24 |
+
import gc
|
| 25 |
+
import json
|
| 26 |
+
import numpy as np
|
| 27 |
+
from tqdm import tqdm
|
| 28 |
+
from collections import OrderedDict
|
| 29 |
+
from dataclasses import dataclass
|
| 30 |
+
|
| 31 |
+
# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
|
| 32 |
+
# DeepSpeed data structures it has to be available in the current python environment.
|
| 33 |
+
from deepspeed.utils import logger
|
| 34 |
+
from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
|
| 35 |
+
FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
|
| 36 |
+
FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
|
| 37 |
+
|
| 38 |
+
|
| 39 |
+
@dataclass
|
| 40 |
+
class zero_model_state:
|
| 41 |
+
buffers: dict()
|
| 42 |
+
param_shapes: dict()
|
| 43 |
+
shared_params: list
|
| 44 |
+
ds_version: int
|
| 45 |
+
frozen_param_shapes: dict()
|
| 46 |
+
frozen_param_fragments: dict()
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
debug = 0
|
| 50 |
+
|
| 51 |
+
# load to cpu
|
| 52 |
+
device = torch.device('cpu')
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
def atoi(text):
|
| 56 |
+
return int(text) if text.isdigit() else text
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
def natural_keys(text):
|
| 60 |
+
'''
|
| 61 |
+
alist.sort(key=natural_keys) sorts in human order
|
| 62 |
+
http://nedbatchelder.com/blog/200712/human_sorting.html
|
| 63 |
+
(See Toothy's implementation in the comments)
|
| 64 |
+
'''
|
| 65 |
+
return [atoi(c) for c in re.split(r'(\d+)', text)]
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
def get_model_state_file(checkpoint_dir, zero_stage):
|
| 69 |
+
if not os.path.isdir(checkpoint_dir):
|
| 70 |
+
raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
|
| 71 |
+
|
| 72 |
+
# there should be only one file
|
| 73 |
+
if zero_stage <= 2:
|
| 74 |
+
file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
|
| 75 |
+
elif zero_stage == 3:
|
| 76 |
+
file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
|
| 77 |
+
|
| 78 |
+
if not os.path.exists(file):
|
| 79 |
+
raise FileNotFoundError(f"can't find model states file at '{file}'")
|
| 80 |
+
|
| 81 |
+
return file
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
def get_checkpoint_files(checkpoint_dir, glob_pattern):
|
| 85 |
+
# XXX: need to test that this simple glob rule works for multi-node setup too
|
| 86 |
+
ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
|
| 87 |
+
|
| 88 |
+
if len(ckpt_files) == 0:
|
| 89 |
+
raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
|
| 90 |
+
|
| 91 |
+
return ckpt_files
|
| 92 |
+
|
| 93 |
+
|
| 94 |
+
def get_optim_files(checkpoint_dir):
|
| 95 |
+
return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
|
| 96 |
+
|
| 97 |
+
|
| 98 |
+
def get_model_state_files(checkpoint_dir):
|
| 99 |
+
return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
def parse_model_states(files):
|
| 103 |
+
zero_model_states = []
|
| 104 |
+
for file in files:
|
| 105 |
+
state_dict = torch.load(file, map_location=device, weights_only=False)
|
| 106 |
+
|
| 107 |
+
if BUFFER_NAMES not in state_dict:
|
| 108 |
+
raise ValueError(f"{file} is not a model state checkpoint")
|
| 109 |
+
buffer_names = state_dict[BUFFER_NAMES]
|
| 110 |
+
if debug:
|
| 111 |
+
print("Found buffers:", buffer_names)
|
| 112 |
+
|
| 113 |
+
# recover just the buffers while restoring them to fp32 if they were saved in fp16
|
| 114 |
+
buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
|
| 115 |
+
param_shapes = state_dict[PARAM_SHAPES]
|
| 116 |
+
|
| 117 |
+
# collect parameters that are included in param_shapes
|
| 118 |
+
param_names = []
|
| 119 |
+
for s in param_shapes:
|
| 120 |
+
for name in s.keys():
|
| 121 |
+
param_names.append(name)
|
| 122 |
+
|
| 123 |
+
# update with frozen parameters
|
| 124 |
+
frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
|
| 125 |
+
if frozen_param_shapes is not None:
|
| 126 |
+
if debug:
|
| 127 |
+
print(f"Found frozen_param_shapes: {frozen_param_shapes}")
|
| 128 |
+
param_names += list(frozen_param_shapes.keys())
|
| 129 |
+
|
| 130 |
+
# handle shared params
|
| 131 |
+
shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
|
| 132 |
+
|
| 133 |
+
ds_version = state_dict.get(DS_VERSION, None)
|
| 134 |
+
|
| 135 |
+
frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
|
| 136 |
+
|
| 137 |
+
z_model_state = zero_model_state(buffers=buffers,
|
| 138 |
+
param_shapes=param_shapes,
|
| 139 |
+
shared_params=shared_params,
|
| 140 |
+
ds_version=ds_version,
|
| 141 |
+
frozen_param_shapes=frozen_param_shapes,
|
| 142 |
+
frozen_param_fragments=frozen_param_fragments)
|
| 143 |
+
zero_model_states.append(z_model_state)
|
| 144 |
+
|
| 145 |
+
return zero_model_states
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
def parse_optim_states(files, ds_checkpoint_dir):
|
| 149 |
+
total_files = len(files)
|
| 150 |
+
state_dicts = []
|
| 151 |
+
for f in tqdm(files, desc='Loading checkpoint shards'):
|
| 152 |
+
state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
|
| 153 |
+
# immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
|
| 154 |
+
# and also handle the case where it was already removed by another helper script
|
| 155 |
+
state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
|
| 156 |
+
state_dicts.append(state_dict)
|
| 157 |
+
|
| 158 |
+
if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]:
|
| 159 |
+
raise ValueError(f"{files[0]} is not a zero checkpoint")
|
| 160 |
+
zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
|
| 161 |
+
world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
|
| 162 |
+
|
| 163 |
+
# For ZeRO-2 each param group can have different partition_count as data parallelism for expert
|
| 164 |
+
# parameters can be different from data parallelism for non-expert parameters. So we can just
|
| 165 |
+
# use the max of the partition_count to get the dp world_size.
|
| 166 |
+
|
| 167 |
+
if type(world_size) is list:
|
| 168 |
+
world_size = max(world_size)
|
| 169 |
+
|
| 170 |
+
if world_size != total_files:
|
| 171 |
+
raise ValueError(
|
| 172 |
+
f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
|
| 173 |
+
"Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
|
| 174 |
+
)
|
| 175 |
+
|
| 176 |
+
# the groups are named differently in each stage
|
| 177 |
+
if zero_stage <= 2:
|
| 178 |
+
fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
|
| 179 |
+
elif zero_stage == 3:
|
| 180 |
+
fp32_groups_key = FP32_FLAT_GROUPS
|
| 181 |
+
else:
|
| 182 |
+
raise ValueError(f"unknown zero stage {zero_stage}")
|
| 183 |
+
|
| 184 |
+
fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
|
| 185 |
+
return zero_stage, world_size, fp32_flat_groups
|
| 186 |
+
|
| 187 |
+
|
| 188 |
+
def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
|
| 189 |
+
"""
|
| 190 |
+
Returns fp32 state_dict reconstructed from ds checkpoint
|
| 191 |
+
|
| 192 |
+
Args:
|
| 193 |
+
- ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
|
| 194 |
+
|
| 195 |
+
"""
|
| 196 |
+
print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
|
| 197 |
+
|
| 198 |
+
optim_files = get_optim_files(ds_checkpoint_dir)
|
| 199 |
+
zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
|
| 200 |
+
print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
|
| 201 |
+
|
| 202 |
+
model_files = get_model_state_files(ds_checkpoint_dir)
|
| 203 |
+
|
| 204 |
+
zero_model_states = parse_model_states(model_files)
|
| 205 |
+
print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
|
| 206 |
+
|
| 207 |
+
if zero_stage <= 2:
|
| 208 |
+
return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 209 |
+
exclude_frozen_parameters)
|
| 210 |
+
elif zero_stage == 3:
|
| 211 |
+
return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 212 |
+
exclude_frozen_parameters)
|
| 213 |
+
|
| 214 |
+
|
| 215 |
+
def _zero2_merge_frozen_params(state_dict, zero_model_states):
|
| 216 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 217 |
+
return
|
| 218 |
+
|
| 219 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 220 |
+
frozen_param_fragments = zero_model_states[0].frozen_param_fragments
|
| 221 |
+
|
| 222 |
+
if debug:
|
| 223 |
+
num_elem = sum(s.numel() for s in frozen_param_shapes.values())
|
| 224 |
+
print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 225 |
+
|
| 226 |
+
wanted_params = len(frozen_param_shapes)
|
| 227 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 228 |
+
avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
|
| 229 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 230 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 231 |
+
|
| 232 |
+
total_params = 0
|
| 233 |
+
total_numel = 0
|
| 234 |
+
for name, shape in frozen_param_shapes.items():
|
| 235 |
+
total_params += 1
|
| 236 |
+
unpartitioned_numel = shape.numel()
|
| 237 |
+
total_numel += unpartitioned_numel
|
| 238 |
+
|
| 239 |
+
state_dict[name] = frozen_param_fragments[name]
|
| 240 |
+
|
| 241 |
+
if debug:
|
| 242 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 243 |
+
|
| 244 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 245 |
+
|
| 246 |
+
|
| 247 |
+
def _has_callable(obj, fn):
|
| 248 |
+
attr = getattr(obj, fn, None)
|
| 249 |
+
return callable(attr)
|
| 250 |
+
|
| 251 |
+
|
| 252 |
+
def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 253 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 254 |
+
|
| 255 |
+
# Reconstruction protocol:
|
| 256 |
+
#
|
| 257 |
+
# XXX: document this
|
| 258 |
+
|
| 259 |
+
if debug:
|
| 260 |
+
for i in range(world_size):
|
| 261 |
+
for j in range(len(fp32_flat_groups[0])):
|
| 262 |
+
print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
|
| 263 |
+
|
| 264 |
+
# XXX: memory usage doubles here (zero2)
|
| 265 |
+
num_param_groups = len(fp32_flat_groups[0])
|
| 266 |
+
merged_single_partition_of_fp32_groups = []
|
| 267 |
+
for i in range(num_param_groups):
|
| 268 |
+
merged_partitions = [sd[i] for sd in fp32_flat_groups]
|
| 269 |
+
full_single_fp32_vector = torch.cat(merged_partitions, 0)
|
| 270 |
+
merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
|
| 271 |
+
avail_numel = sum(
|
| 272 |
+
[full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
|
| 273 |
+
|
| 274 |
+
if debug:
|
| 275 |
+
wanted_params = sum([len(shapes) for shapes in param_shapes])
|
| 276 |
+
wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
|
| 277 |
+
# not asserting if there is a mismatch due to possible padding
|
| 278 |
+
print(f"Have {avail_numel} numels to process.")
|
| 279 |
+
print(f"Need {wanted_numel} numels in {wanted_params} params.")
|
| 280 |
+
|
| 281 |
+
# params
|
| 282 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 283 |
+
# out-of-core computing solution
|
| 284 |
+
total_numel = 0
|
| 285 |
+
total_params = 0
|
| 286 |
+
for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
|
| 287 |
+
offset = 0
|
| 288 |
+
avail_numel = full_single_fp32_vector.numel()
|
| 289 |
+
for name, shape in shapes.items():
|
| 290 |
+
|
| 291 |
+
unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
|
| 292 |
+
total_numel += unpartitioned_numel
|
| 293 |
+
total_params += 1
|
| 294 |
+
|
| 295 |
+
if debug:
|
| 296 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 297 |
+
state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
|
| 298 |
+
offset += unpartitioned_numel
|
| 299 |
+
|
| 300 |
+
# Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
|
| 301 |
+
# avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
|
| 302 |
+
# paddings performed in the code it's almost impossible to predict the exact numbers w/o the
|
| 303 |
+
# live optimizer object, so we are checking that the numbers are within the right range
|
| 304 |
+
align_to = 2 * world_size
|
| 305 |
+
|
| 306 |
+
def zero2_align(x):
|
| 307 |
+
return align_to * math.ceil(x / align_to)
|
| 308 |
+
|
| 309 |
+
if debug:
|
| 310 |
+
print(f"original offset={offset}, avail_numel={avail_numel}")
|
| 311 |
+
|
| 312 |
+
offset = zero2_align(offset)
|
| 313 |
+
avail_numel = zero2_align(avail_numel)
|
| 314 |
+
|
| 315 |
+
if debug:
|
| 316 |
+
print(f"aligned offset={offset}, avail_numel={avail_numel}")
|
| 317 |
+
|
| 318 |
+
# Sanity check
|
| 319 |
+
if offset != avail_numel:
|
| 320 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 321 |
+
|
| 322 |
+
print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
|
| 323 |
+
|
| 324 |
+
|
| 325 |
+
def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 326 |
+
exclude_frozen_parameters):
|
| 327 |
+
state_dict = OrderedDict()
|
| 328 |
+
|
| 329 |
+
# buffers
|
| 330 |
+
buffers = zero_model_states[0].buffers
|
| 331 |
+
state_dict.update(buffers)
|
| 332 |
+
if debug:
|
| 333 |
+
print(f"added {len(buffers)} buffers")
|
| 334 |
+
|
| 335 |
+
if not exclude_frozen_parameters:
|
| 336 |
+
_zero2_merge_frozen_params(state_dict, zero_model_states)
|
| 337 |
+
|
| 338 |
+
_zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 339 |
+
|
| 340 |
+
# recover shared parameters
|
| 341 |
+
for pair in zero_model_states[0].shared_params:
|
| 342 |
+
if pair[1] in state_dict:
|
| 343 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 344 |
+
|
| 345 |
+
return state_dict
|
| 346 |
+
|
| 347 |
+
|
| 348 |
+
def zero3_partitioned_param_info(unpartitioned_numel, world_size):
|
| 349 |
+
remainder = unpartitioned_numel % world_size
|
| 350 |
+
padding_numel = (world_size - remainder) if remainder else 0
|
| 351 |
+
partitioned_numel = math.ceil(unpartitioned_numel / world_size)
|
| 352 |
+
return partitioned_numel, padding_numel
|
| 353 |
+
|
| 354 |
+
|
| 355 |
+
def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
|
| 356 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 357 |
+
return
|
| 358 |
+
|
| 359 |
+
if debug:
|
| 360 |
+
for i in range(world_size):
|
| 361 |
+
num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
|
| 362 |
+
print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 363 |
+
|
| 364 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 365 |
+
wanted_params = len(frozen_param_shapes)
|
| 366 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 367 |
+
avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
|
| 368 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 369 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 370 |
+
|
| 371 |
+
total_params = 0
|
| 372 |
+
total_numel = 0
|
| 373 |
+
for name, shape in zero_model_states[0].frozen_param_shapes.items():
|
| 374 |
+
total_params += 1
|
| 375 |
+
unpartitioned_numel = shape.numel()
|
| 376 |
+
total_numel += unpartitioned_numel
|
| 377 |
+
|
| 378 |
+
param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
|
| 379 |
+
state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 380 |
+
|
| 381 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 382 |
+
|
| 383 |
+
if debug:
|
| 384 |
+
print(
|
| 385 |
+
f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 386 |
+
)
|
| 387 |
+
|
| 388 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 389 |
+
|
| 390 |
+
|
| 391 |
+
class GatheredTensor:
|
| 392 |
+
"""
|
| 393 |
+
A pseudo tensor that collects partitioned weights.
|
| 394 |
+
It is more memory efficient when there are multiple groups.
|
| 395 |
+
"""
|
| 396 |
+
|
| 397 |
+
def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
|
| 398 |
+
self.flat_groups = flat_groups
|
| 399 |
+
self.flat_groups_offset = flat_groups_offset
|
| 400 |
+
self.offset = offset
|
| 401 |
+
self.partitioned_numel = partitioned_numel
|
| 402 |
+
self.shape = shape
|
| 403 |
+
self.dtype = self.flat_groups[0][0].dtype
|
| 404 |
+
|
| 405 |
+
def contiguous(self):
|
| 406 |
+
"""
|
| 407 |
+
Merge partitioned weights from flat_groups into a single tensor.
|
| 408 |
+
"""
|
| 409 |
+
end_idx = self.offset + self.partitioned_numel
|
| 410 |
+
world_size = len(self.flat_groups)
|
| 411 |
+
pad_flat_param_chunks = []
|
| 412 |
+
|
| 413 |
+
for rank_i in range(world_size):
|
| 414 |
+
# for each rank, we need to collect weights from related group/groups
|
| 415 |
+
flat_groups_at_rank_i = self.flat_groups[rank_i]
|
| 416 |
+
start_group_id = None
|
| 417 |
+
end_group_id = None
|
| 418 |
+
for group_id in range(len(self.flat_groups_offset)):
|
| 419 |
+
if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
|
| 420 |
+
start_group_id = group_id
|
| 421 |
+
if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
|
| 422 |
+
end_group_id = group_id
|
| 423 |
+
break
|
| 424 |
+
# collect weights from related group/groups
|
| 425 |
+
for group_id in range(start_group_id, end_group_id + 1):
|
| 426 |
+
flat_tensor = flat_groups_at_rank_i[group_id]
|
| 427 |
+
start_offset = self.offset - self.flat_groups_offset[group_id]
|
| 428 |
+
end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
|
| 429 |
+
pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
|
| 430 |
+
|
| 431 |
+
# collect weights from all ranks
|
| 432 |
+
pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
|
| 433 |
+
param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
|
| 434 |
+
return param
|
| 435 |
+
|
| 436 |
+
|
| 437 |
+
def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 438 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 439 |
+
avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
|
| 440 |
+
|
| 441 |
+
# Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
|
| 442 |
+
# param, re-consolidating each param, while dealing with padding if any
|
| 443 |
+
|
| 444 |
+
# merge list of dicts, preserving order
|
| 445 |
+
param_shapes = {k: v for d in param_shapes for k, v in d.items()}
|
| 446 |
+
|
| 447 |
+
if debug:
|
| 448 |
+
for i in range(world_size):
|
| 449 |
+
print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
|
| 450 |
+
|
| 451 |
+
wanted_params = len(param_shapes)
|
| 452 |
+
wanted_numel = sum(shape.numel() for shape in param_shapes.values())
|
| 453 |
+
# not asserting if there is a mismatch due to possible padding
|
| 454 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 455 |
+
print(f"Trainable params: Have {avail_numel} numels to process.")
|
| 456 |
+
print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
|
| 457 |
+
|
| 458 |
+
# params
|
| 459 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 460 |
+
# out-of-core computing solution
|
| 461 |
+
offset = 0
|
| 462 |
+
total_numel = 0
|
| 463 |
+
total_params = 0
|
| 464 |
+
flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
|
| 465 |
+
for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
|
| 466 |
+
unpartitioned_numel = shape.numel()
|
| 467 |
+
total_numel += unpartitioned_numel
|
| 468 |
+
total_params += 1
|
| 469 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 470 |
+
|
| 471 |
+
if debug:
|
| 472 |
+
print(
|
| 473 |
+
f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 474 |
+
)
|
| 475 |
+
|
| 476 |
+
# memory efficient tensor
|
| 477 |
+
tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
|
| 478 |
+
state_dict[name] = tensor
|
| 479 |
+
offset += partitioned_numel
|
| 480 |
+
|
| 481 |
+
offset *= world_size
|
| 482 |
+
|
| 483 |
+
# Sanity check
|
| 484 |
+
if offset != avail_numel:
|
| 485 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 486 |
+
|
| 487 |
+
print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
|
| 488 |
+
|
| 489 |
+
|
| 490 |
+
def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 491 |
+
exclude_frozen_parameters):
|
| 492 |
+
state_dict = OrderedDict()
|
| 493 |
+
|
| 494 |
+
# buffers
|
| 495 |
+
buffers = zero_model_states[0].buffers
|
| 496 |
+
state_dict.update(buffers)
|
| 497 |
+
if debug:
|
| 498 |
+
print(f"added {len(buffers)} buffers")
|
| 499 |
+
|
| 500 |
+
if not exclude_frozen_parameters:
|
| 501 |
+
_zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
|
| 502 |
+
|
| 503 |
+
_zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 504 |
+
|
| 505 |
+
# recover shared parameters
|
| 506 |
+
for pair in zero_model_states[0].shared_params:
|
| 507 |
+
if pair[1] in state_dict:
|
| 508 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 509 |
+
|
| 510 |
+
return state_dict
|
| 511 |
+
|
| 512 |
+
|
| 513 |
+
def to_torch_tensor(state_dict, return_empty_tensor=False):
|
| 514 |
+
"""
|
| 515 |
+
Convert state_dict of GatheredTensor to torch tensor
|
| 516 |
+
"""
|
| 517 |
+
torch_state_dict = {}
|
| 518 |
+
converted_tensors = {}
|
| 519 |
+
for name, tensor in state_dict.items():
|
| 520 |
+
tensor_id = id(tensor)
|
| 521 |
+
if tensor_id in converted_tensors: # shared tensors
|
| 522 |
+
shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
|
| 523 |
+
torch_state_dict[name] = shared_tensor
|
| 524 |
+
else:
|
| 525 |
+
converted_tensors[tensor_id] = name
|
| 526 |
+
if return_empty_tensor:
|
| 527 |
+
torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
|
| 528 |
+
else:
|
| 529 |
+
torch_state_dict[name] = tensor.contiguous()
|
| 530 |
+
return torch_state_dict
|
| 531 |
+
|
| 532 |
+
|
| 533 |
+
def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 534 |
+
tag=None,
|
| 535 |
+
exclude_frozen_parameters=False,
|
| 536 |
+
lazy_mode=False):
|
| 537 |
+
"""
|
| 538 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
|
| 539 |
+
``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
|
| 540 |
+
via a model hub.
|
| 541 |
+
|
| 542 |
+
Args:
|
| 543 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder
|
| 544 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
|
| 545 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 546 |
+
- ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
|
| 547 |
+
Convert the pesduo tensor to torch tensor by ``.contiguous()``
|
| 548 |
+
|
| 549 |
+
Returns:
|
| 550 |
+
- pytorch ``state_dict``
|
| 551 |
+
|
| 552 |
+
A typical usage might be ::
|
| 553 |
+
|
| 554 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 555 |
+
# do the training and checkpoint saving
|
| 556 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
|
| 557 |
+
model = model.cpu() # move to cpu
|
| 558 |
+
model.load_state_dict(state_dict)
|
| 559 |
+
# submit to model hub or save the model to share with others
|
| 560 |
+
|
| 561 |
+
In this example the ``model`` will no longer be usable in the deepspeed context of the same
|
| 562 |
+
application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 563 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 564 |
+
|
| 565 |
+
If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
|
| 566 |
+
|
| 567 |
+
Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
|
| 568 |
+
You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
|
| 569 |
+
the checkpoint. Or you can load state_dict in lazy mode ::
|
| 570 |
+
|
| 571 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 572 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
|
| 573 |
+
for name, lazy_tensor in state_dict.item():
|
| 574 |
+
tensor = lazy_tensor.contiguous() # to cpu
|
| 575 |
+
print(name, tensor)
|
| 576 |
+
# del tensor to release memory if it no longer in use
|
| 577 |
+
"""
|
| 578 |
+
if tag is None:
|
| 579 |
+
latest_path = os.path.join(checkpoint_dir, 'latest')
|
| 580 |
+
if os.path.isfile(latest_path):
|
| 581 |
+
with open(latest_path, 'r') as fd:
|
| 582 |
+
tag = fd.read().strip()
|
| 583 |
+
else:
|
| 584 |
+
raise ValueError(f"Unable to find 'latest' file at {latest_path}")
|
| 585 |
+
|
| 586 |
+
ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
|
| 587 |
+
|
| 588 |
+
if not os.path.isdir(ds_checkpoint_dir):
|
| 589 |
+
raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
|
| 590 |
+
|
| 591 |
+
state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
|
| 592 |
+
if lazy_mode:
|
| 593 |
+
return state_dict
|
| 594 |
+
else:
|
| 595 |
+
return to_torch_tensor(state_dict)
|
| 596 |
+
|
| 597 |
+
|
| 598 |
+
def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
|
| 599 |
+
output_dir,
|
| 600 |
+
max_shard_size="5GB",
|
| 601 |
+
safe_serialization=False,
|
| 602 |
+
tag=None,
|
| 603 |
+
exclude_frozen_parameters=False):
|
| 604 |
+
"""
|
| 605 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
|
| 606 |
+
loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
|
| 607 |
+
|
| 608 |
+
Args:
|
| 609 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 610 |
+
- ``output_dir``: directory to the pytorch fp32 state_dict output files
|
| 611 |
+
- ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
|
| 612 |
+
- ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
|
| 613 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 614 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 615 |
+
"""
|
| 616 |
+
|
| 617 |
+
# Dependency pre-check
|
| 618 |
+
if safe_serialization:
|
| 619 |
+
try:
|
| 620 |
+
from safetensors.torch import save_file
|
| 621 |
+
except ImportError:
|
| 622 |
+
print('If you want to use `safe_serialization`, please `pip install safetensors`')
|
| 623 |
+
raise
|
| 624 |
+
if max_shard_size is not None:
|
| 625 |
+
try:
|
| 626 |
+
from huggingface_hub import split_torch_state_dict_into_shards
|
| 627 |
+
except ImportError:
|
| 628 |
+
print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
|
| 629 |
+
raise
|
| 630 |
+
|
| 631 |
+
# Convert zero checkpoint to state_dict
|
| 632 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 633 |
+
tag,
|
| 634 |
+
exclude_frozen_parameters,
|
| 635 |
+
lazy_mode=True)
|
| 636 |
+
|
| 637 |
+
# Shard the model if it is too big.
|
| 638 |
+
weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
|
| 639 |
+
if max_shard_size is not None:
|
| 640 |
+
filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
|
| 641 |
+
# an memory-efficient approach for sharding
|
| 642 |
+
empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
|
| 643 |
+
state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
|
| 644 |
+
filename_pattern=filename_pattern,
|
| 645 |
+
max_shard_size=max_shard_size)
|
| 646 |
+
else:
|
| 647 |
+
from collections import namedtuple
|
| 648 |
+
StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
|
| 649 |
+
state_dict_split = StateDictSplit(is_sharded=False,
|
| 650 |
+
filename_to_tensors={weights_name: list(state_dict.keys())})
|
| 651 |
+
|
| 652 |
+
# Save the model by shard
|
| 653 |
+
os.makedirs(output_dir, exist_ok=True)
|
| 654 |
+
filename_to_tensors = state_dict_split.filename_to_tensors.items()
|
| 655 |
+
for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
|
| 656 |
+
shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
|
| 657 |
+
shard_state_dict = to_torch_tensor(shard_state_dict)
|
| 658 |
+
output_path = os.path.join(output_dir, shard_file)
|
| 659 |
+
if safe_serialization:
|
| 660 |
+
save_file(shard_state_dict, output_path, metadata={"format": "pt"})
|
| 661 |
+
else:
|
| 662 |
+
torch.save(shard_state_dict, output_path)
|
| 663 |
+
# release the memory of current shard
|
| 664 |
+
for tensor_name in list(shard_state_dict.keys()):
|
| 665 |
+
del state_dict[tensor_name]
|
| 666 |
+
del shard_state_dict[tensor_name]
|
| 667 |
+
del shard_state_dict
|
| 668 |
+
gc.collect()
|
| 669 |
+
|
| 670 |
+
# Save index if sharded
|
| 671 |
+
if state_dict_split.is_sharded:
|
| 672 |
+
index = {
|
| 673 |
+
"metadata": state_dict_split.metadata,
|
| 674 |
+
"weight_map": state_dict_split.tensor_to_filename,
|
| 675 |
+
}
|
| 676 |
+
save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
|
| 677 |
+
save_index_file = os.path.join(output_dir, save_index_file)
|
| 678 |
+
with open(save_index_file, "w", encoding="utf-8") as f:
|
| 679 |
+
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
|
| 680 |
+
f.write(content)
|
| 681 |
+
|
| 682 |
+
|
| 683 |
+
def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
|
| 684 |
+
"""
|
| 685 |
+
1. Put the provided model to cpu
|
| 686 |
+
2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
|
| 687 |
+
3. Load it into the provided model
|
| 688 |
+
|
| 689 |
+
Args:
|
| 690 |
+
- ``model``: the model object to update
|
| 691 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 692 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 693 |
+
|
| 694 |
+
Returns:
|
| 695 |
+
- ``model`: modified model
|
| 696 |
+
|
| 697 |
+
Make sure you have plenty of CPU memory available before you call this function. If you don't
|
| 698 |
+
have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
|
| 699 |
+
conveniently placed for you in the checkpoint folder.
|
| 700 |
+
|
| 701 |
+
A typical usage might be ::
|
| 702 |
+
|
| 703 |
+
from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
|
| 704 |
+
model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
|
| 705 |
+
# submit to model hub or save the model to share with others
|
| 706 |
+
|
| 707 |
+
Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
|
| 708 |
+
of the same application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 709 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 710 |
+
|
| 711 |
+
"""
|
| 712 |
+
logger.info("Extracting fp32 weights")
|
| 713 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
|
| 714 |
+
|
| 715 |
+
logger.info("Overwriting model with fp32 weights")
|
| 716 |
+
model = model.cpu()
|
| 717 |
+
model.load_state_dict(state_dict, strict=False)
|
| 718 |
+
|
| 719 |
+
return model
|
| 720 |
+
|
| 721 |
+
|
| 722 |
+
if __name__ == "__main__":
|
| 723 |
+
parser = argparse.ArgumentParser()
|
| 724 |
+
parser.add_argument("checkpoint_dir",
|
| 725 |
+
type=str,
|
| 726 |
+
help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
|
| 727 |
+
parser.add_argument("output_dir",
|
| 728 |
+
type=str,
|
| 729 |
+
help="directory to the pytorch fp32 state_dict output files"
|
| 730 |
+
"(e.g. path/checkpoint-12-output/)")
|
| 731 |
+
parser.add_argument(
|
| 732 |
+
"--max_shard_size",
|
| 733 |
+
type=str,
|
| 734 |
+
default="5GB",
|
| 735 |
+
help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
|
| 736 |
+
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
|
| 737 |
+
"We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
|
| 738 |
+
"without CPU OOM issues.")
|
| 739 |
+
parser.add_argument(
|
| 740 |
+
"--safe_serialization",
|
| 741 |
+
default=False,
|
| 742 |
+
action='store_true',
|
| 743 |
+
help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
|
| 744 |
+
parser.add_argument("-t",
|
| 745 |
+
"--tag",
|
| 746 |
+
type=str,
|
| 747 |
+
default=None,
|
| 748 |
+
help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
|
| 749 |
+
parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
|
| 750 |
+
parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
|
| 751 |
+
args = parser.parse_args()
|
| 752 |
+
|
| 753 |
+
debug = args.debug
|
| 754 |
+
|
| 755 |
+
convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
|
| 756 |
+
args.output_dir,
|
| 757 |
+
max_shard_size=args.max_shard_size,
|
| 758 |
+
safe_serialization=args.safe_serialization,
|
| 759 |
+
tag=args.tag,
|
| 760 |
+
exclude_frozen_parameters=args.exclude_frozen_parameters)
|