End of training
Browse files- README.md +56 -0
- adapter_config.json +34 -0
- adapter_model.safetensors +3 -0
- trainer_state.json +119 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
base_model: yuiseki/karasu-aya-ja-v0.1
|
6 |
+
model-index:
|
7 |
+
- name: karasu-aya-ft-onlyqa-adapter
|
8 |
+
results: []
|
9 |
+
---
|
10 |
+
|
11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
12 |
+
should probably proofread and complete it, then remove this comment. -->
|
13 |
+
|
14 |
+
# karasu-aya-ft-onlyqa-adapter
|
15 |
+
|
16 |
+
This model is a fine-tuned version of [yuiseki/karasu-aya-ja-v0.1](https://huggingface.co/yuiseki/karasu-aya-ja-v0.1) on the None dataset.
|
17 |
+
|
18 |
+
## Model description
|
19 |
+
|
20 |
+
More information needed
|
21 |
+
|
22 |
+
## Intended uses & limitations
|
23 |
+
|
24 |
+
More information needed
|
25 |
+
|
26 |
+
## Training and evaluation data
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Training procedure
|
31 |
+
|
32 |
+
### Training hyperparameters
|
33 |
+
|
34 |
+
The following hyperparameters were used during training:
|
35 |
+
- learning_rate: 0.0001
|
36 |
+
- train_batch_size: 1
|
37 |
+
- eval_batch_size: 8
|
38 |
+
- seed: 42
|
39 |
+
- gradient_accumulation_steps: 2
|
40 |
+
- total_train_batch_size: 2
|
41 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
42 |
+
- lr_scheduler_type: linear
|
43 |
+
- lr_scheduler_warmup_steps: 5
|
44 |
+
- num_epochs: 1
|
45 |
+
|
46 |
+
### Training results
|
47 |
+
|
48 |
+
|
49 |
+
|
50 |
+
### Framework versions
|
51 |
+
|
52 |
+
- PEFT 0.11.1
|
53 |
+
- Transformers 4.41.1
|
54 |
+
- Pytorch 2.3.0
|
55 |
+
- Datasets 2.20.0
|
56 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": null,
|
4 |
+
"base_model_name_or_path": "yuiseki/karasu-aya-ja-v0.1",
|
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": 32,
|
14 |
+
"lora_dropout": 0.05,
|
15 |
+
"megatron_config": null,
|
16 |
+
"megatron_core": "megatron.core",
|
17 |
+
"modules_to_save": null,
|
18 |
+
"peft_type": "LORA",
|
19 |
+
"r": 32,
|
20 |
+
"rank_pattern": {},
|
21 |
+
"revision": null,
|
22 |
+
"target_modules": [
|
23 |
+
"mlp.gate_proj",
|
24 |
+
"self_attn.v_proj",
|
25 |
+
"mlp.up_proj",
|
26 |
+
"mlp.down_proj",
|
27 |
+
"self_attn.k_proj",
|
28 |
+
"self_attn.q_proj",
|
29 |
+
"self_attn.o_proj"
|
30 |
+
],
|
31 |
+
"task_type": "CAUSAL_LM",
|
32 |
+
"use_dora": false,
|
33 |
+
"use_rslora": false
|
34 |
+
}
|
adapter_model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1763fa29d3aaed609a9761ccc62f3364e46082179eaf8a864774c41c2931a896
|
3 |
+
size 100966336
|
trainer_state.json
ADDED
@@ -0,0 +1,119 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.9911504424778761,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 56,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.08849557522123894,
|
13 |
+
"grad_norm": 0.6312227845191956,
|
14 |
+
"learning_rate": 0.0001,
|
15 |
+
"loss": 1.6685,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.17699115044247787,
|
20 |
+
"grad_norm": 0.35548320412635803,
|
21 |
+
"learning_rate": 9.019607843137255e-05,
|
22 |
+
"loss": 1.4885,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.26548672566371684,
|
27 |
+
"grad_norm": 0.612781286239624,
|
28 |
+
"learning_rate": 8.039215686274511e-05,
|
29 |
+
"loss": 1.3795,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.35398230088495575,
|
34 |
+
"grad_norm": 0.7122682929039001,
|
35 |
+
"learning_rate": 7.058823529411765e-05,
|
36 |
+
"loss": 1.2164,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.4424778761061947,
|
41 |
+
"grad_norm": 0.6391258835792542,
|
42 |
+
"learning_rate": 6.078431372549019e-05,
|
43 |
+
"loss": 0.9956,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.5309734513274337,
|
48 |
+
"grad_norm": 0.4501003324985504,
|
49 |
+
"learning_rate": 5.0980392156862745e-05,
|
50 |
+
"loss": 1.0388,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.6194690265486725,
|
55 |
+
"grad_norm": 0.35456234216690063,
|
56 |
+
"learning_rate": 4.11764705882353e-05,
|
57 |
+
"loss": 1.014,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.7079646017699115,
|
62 |
+
"grad_norm": 0.4066126048564911,
|
63 |
+
"learning_rate": 3.137254901960784e-05,
|
64 |
+
"loss": 0.9312,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.7964601769911505,
|
69 |
+
"grad_norm": 0.4473574161529541,
|
70 |
+
"learning_rate": 2.1568627450980395e-05,
|
71 |
+
"loss": 0.8882,
|
72 |
+
"step": 45
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.8849557522123894,
|
76 |
+
"grad_norm": 0.42478886246681213,
|
77 |
+
"learning_rate": 1.1764705882352942e-05,
|
78 |
+
"loss": 0.9296,
|
79 |
+
"step": 50
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.9734513274336283,
|
83 |
+
"grad_norm": 0.493737131357193,
|
84 |
+
"learning_rate": 1.96078431372549e-06,
|
85 |
+
"loss": 0.8824,
|
86 |
+
"step": 55
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.9911504424778761,
|
90 |
+
"step": 56,
|
91 |
+
"total_flos": 214057638850560.0,
|
92 |
+
"train_loss": 1.126003892293998,
|
93 |
+
"train_runtime": 11094.3613,
|
94 |
+
"train_samples_per_second": 0.01,
|
95 |
+
"train_steps_per_second": 0.005
|
96 |
+
}
|
97 |
+
],
|
98 |
+
"logging_steps": 5,
|
99 |
+
"max_steps": 56,
|
100 |
+
"num_input_tokens_seen": 0,
|
101 |
+
"num_train_epochs": 1,
|
102 |
+
"save_steps": 500,
|
103 |
+
"stateful_callbacks": {
|
104 |
+
"TrainerControl": {
|
105 |
+
"args": {
|
106 |
+
"should_epoch_stop": false,
|
107 |
+
"should_evaluate": false,
|
108 |
+
"should_log": false,
|
109 |
+
"should_save": false,
|
110 |
+
"should_training_stop": false
|
111 |
+
},
|
112 |
+
"attributes": {}
|
113 |
+
}
|
114 |
+
},
|
115 |
+
"total_flos": 214057638850560.0,
|
116 |
+
"train_batch_size": 1,
|
117 |
+
"trial_name": null,
|
118 |
+
"trial_params": null
|
119 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:04ae0f9527e3c5c8681cbfa53a8155a7dfb4ab093deede0e22c7b0c8ffe7e893
|
3 |
+
size 5112
|