Upload 5 files
Browse files- adapter_config.json +34 -0
- added_tokens.json +24 -0
- all_results.json +9 -0
- infer.yaml +4 -0
- llamaboard_config.yaml +67 -0
adapter_config.json
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": null,
|
4 |
+
"base_model_name_or_path": "/mnt/w2/lzq/nex_LLM/weights/Qwen2.5-3B-Instruct",
|
5 |
+
"bias": "none",
|
6 |
+
"fan_in_fan_out": false,
|
7 |
+
"inference_mode": true,
|
8 |
+
"init_lora_weights": true,
|
9 |
+
"layer_replication": null,
|
10 |
+
"layers_pattern": null,
|
11 |
+
"layers_to_transform": null,
|
12 |
+
"loftq_config": {},
|
13 |
+
"lora_alpha": 16,
|
14 |
+
"lora_dropout": 0,
|
15 |
+
"megatron_config": null,
|
16 |
+
"megatron_core": "megatron.core",
|
17 |
+
"modules_to_save": null,
|
18 |
+
"peft_type": "LORA",
|
19 |
+
"r": 8,
|
20 |
+
"rank_pattern": {},
|
21 |
+
"revision": null,
|
22 |
+
"target_modules": [
|
23 |
+
"gate_proj",
|
24 |
+
"up_proj",
|
25 |
+
"k_proj",
|
26 |
+
"o_proj",
|
27 |
+
"v_proj",
|
28 |
+
"q_proj",
|
29 |
+
"down_proj"
|
30 |
+
],
|
31 |
+
"task_type": "CAUSAL_LM",
|
32 |
+
"use_dora": false,
|
33 |
+
"use_rslora": false
|
34 |
+
}
|
added_tokens.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</tool_call>": 151658,
|
3 |
+
"<tool_call>": 151657,
|
4 |
+
"<|box_end|>": 151649,
|
5 |
+
"<|box_start|>": 151648,
|
6 |
+
"<|endoftext|>": 151643,
|
7 |
+
"<|file_sep|>": 151664,
|
8 |
+
"<|fim_middle|>": 151660,
|
9 |
+
"<|fim_pad|>": 151662,
|
10 |
+
"<|fim_prefix|>": 151659,
|
11 |
+
"<|fim_suffix|>": 151661,
|
12 |
+
"<|im_end|>": 151645,
|
13 |
+
"<|im_start|>": 151644,
|
14 |
+
"<|image_pad|>": 151655,
|
15 |
+
"<|object_ref_end|>": 151647,
|
16 |
+
"<|object_ref_start|>": 151646,
|
17 |
+
"<|quad_end|>": 151651,
|
18 |
+
"<|quad_start|>": 151650,
|
19 |
+
"<|repo_name|>": 151663,
|
20 |
+
"<|video_pad|>": 151656,
|
21 |
+
"<|vision_end|>": 151653,
|
22 |
+
"<|vision_pad|>": 151654,
|
23 |
+
"<|vision_start|>": 151652
|
24 |
+
}
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 44.44444444444444,
|
3 |
+
"num_input_tokens_seen": 349136,
|
4 |
+
"total_flos": 5843993170542592.0,
|
5 |
+
"train_loss": 0.6807288352648417,
|
6 |
+
"train_runtime": 762.8508,
|
7 |
+
"train_samples_per_second": 7.079,
|
8 |
+
"train_steps_per_second": 0.197
|
9 |
+
}
|
infer.yaml
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
model_name_or_path: /mnt/w2/lzq/nex_LLM/weights/Qwen2.5-3B-Instruct
|
2 |
+
template: qwen
|
3 |
+
infer_backend: vllm
|
4 |
+
vllm_enforce_eager: true
|
llamaboard_config.yaml
ADDED
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
top.booster: auto
|
2 |
+
top.checkpoint_path:
|
3 |
+
- train_2024-10-31-11-14-25
|
4 |
+
top.finetuning_type: lora
|
5 |
+
top.model_name: Qwen2.5-3B-Instruct
|
6 |
+
top.quantization_bit: none
|
7 |
+
top.quantization_method: bitsandbytes
|
8 |
+
top.rope_scaling: none
|
9 |
+
top.template: qwen
|
10 |
+
train.additional_target: ''
|
11 |
+
train.badam_mode: layer
|
12 |
+
train.badam_switch_interval: 50
|
13 |
+
train.badam_switch_mode: ascending
|
14 |
+
train.badam_update_ratio: 0.05
|
15 |
+
train.batch_size: 2
|
16 |
+
train.compute_type: fp16
|
17 |
+
train.create_new_adapter: false
|
18 |
+
train.cutoff_len: 1024
|
19 |
+
train.dataset:
|
20 |
+
- self_cognition
|
21 |
+
train.dataset_dir: data
|
22 |
+
train.ds_offload: false
|
23 |
+
train.ds_stage: none
|
24 |
+
train.freeze_extra_modules: ''
|
25 |
+
train.freeze_trainable_layers: 2
|
26 |
+
train.freeze_trainable_modules: all
|
27 |
+
train.galore_rank: 16
|
28 |
+
train.galore_scale: 0.25
|
29 |
+
train.galore_target: all
|
30 |
+
train.galore_update_interval: 200
|
31 |
+
train.gradient_accumulation_steps: 8
|
32 |
+
train.learning_rate: 5e-5
|
33 |
+
train.logging_steps: 5
|
34 |
+
train.lora_alpha: 16
|
35 |
+
train.lora_dropout: 0
|
36 |
+
train.lora_rank: 8
|
37 |
+
train.lora_target: all
|
38 |
+
train.loraplus_lr_ratio: 0
|
39 |
+
train.lr_scheduler_type: cosine
|
40 |
+
train.mask_history: false
|
41 |
+
train.max_grad_norm: '1.0'
|
42 |
+
train.max_samples: '100000'
|
43 |
+
train.neat_packing: false
|
44 |
+
train.neftune_alpha: 0
|
45 |
+
train.num_train_epochs: '100'
|
46 |
+
train.optim: adamw_torch
|
47 |
+
train.packing: false
|
48 |
+
train.ppo_score_norm: false
|
49 |
+
train.ppo_whiten_rewards: false
|
50 |
+
train.pref_beta: 0.1
|
51 |
+
train.pref_ftx: 0
|
52 |
+
train.pref_loss: sigmoid
|
53 |
+
train.report_to: false
|
54 |
+
train.resize_vocab: false
|
55 |
+
train.reward_model: null
|
56 |
+
train.save_steps: 100
|
57 |
+
train.shift_attn: false
|
58 |
+
train.train_on_prompt: false
|
59 |
+
train.training_stage: Supervised Fine-Tuning
|
60 |
+
train.use_badam: false
|
61 |
+
train.use_dora: false
|
62 |
+
train.use_galore: false
|
63 |
+
train.use_llama_pro: false
|
64 |
+
train.use_pissa: false
|
65 |
+
train.use_rslora: false
|
66 |
+
train.val_size: 0
|
67 |
+
train.warmup_steps: 0
|