Upload 23 files
Browse files- lora-alpaca/adapter_config.json +18 -0
- lora-alpaca/adapter_model.bin +3 -0
- lora-alpaca/checkpoint-1000/optimizer.pt +3 -0
- lora-alpaca/checkpoint-1000/pytorch_model.bin +3 -0
- lora-alpaca/checkpoint-1000/rng_state.pth +3 -0
- lora-alpaca/checkpoint-1000/scaler.pt +3 -0
- lora-alpaca/checkpoint-1000/scheduler.pt +3 -0
- lora-alpaca/checkpoint-1000/trainer_state.json +656 -0
- lora-alpaca/checkpoint-1000/training_args.bin +3 -0
- lora-alpaca/checkpoint-1200/optimizer.pt +3 -0
- lora-alpaca/checkpoint-1200/pytorch_model.bin +3 -0
- lora-alpaca/checkpoint-1200/rng_state.pth +3 -0
- lora-alpaca/checkpoint-1200/scaler.pt +3 -0
- lora-alpaca/checkpoint-1200/scheduler.pt +3 -0
- lora-alpaca/checkpoint-1200/trainer_state.json +784 -0
- lora-alpaca/checkpoint-1200/training_args.bin +3 -0
- lora-alpaca/checkpoint-1400/optimizer.pt +3 -0
- lora-alpaca/checkpoint-1400/pytorch_model.bin +3 -0
- lora-alpaca/checkpoint-1400/rng_state.pth +3 -0
- lora-alpaca/checkpoint-1400/scaler.pt +3 -0
- lora-alpaca/checkpoint-1400/scheduler.pt +3 -0
- lora-alpaca/checkpoint-1400/trainer_state.json +912 -0
- lora-alpaca/checkpoint-1400/training_args.bin +3 -0
lora-alpaca/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "decapoda-research/llama-7b-hf",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"enable_lora": null,
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"lora_alpha": 16,
|
| 8 |
+
"lora_dropout": 0.05,
|
| 9 |
+
"merge_weights": false,
|
| 10 |
+
"modules_to_save": null,
|
| 11 |
+
"peft_type": "LORA",
|
| 12 |
+
"r": 8,
|
| 13 |
+
"target_modules": [
|
| 14 |
+
"q_proj",
|
| 15 |
+
"v_proj"
|
| 16 |
+
],
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
lora-alpaca/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:069b50d834100806cf46bddd3ced0573fc39bf958dd699ef0544fd8562d10e5c
|
| 3 |
+
size 16822989
|
lora-alpaca/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84de79920309c19125174dfd6f66d92cff868fd3aabe7e3c90b606a9d250f57b
|
| 3 |
+
size 33629893
|
lora-alpaca/checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:337df0b61026374e5092237fc8cfc397ed81af4f617325aa27b0a9aecff64415
|
| 3 |
+
size 16822989
|
lora-alpaca/checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32d66b7b6fc312e3cb6d1e612a05ff8630875d5559c78e8d151c1aa97988ad44
|
| 3 |
+
size 14575
|
lora-alpaca/checkpoint-1000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68cff80b680ddf6e7abbef98b5f336b97f9b5963e2209307f639383870e8cc71
|
| 3 |
+
size 557
|
lora-alpaca/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dee6f5ebda8cadcb8b98dd5fe5df932a2cf249706a60d0cef298c621c5255bf
|
| 3 |
+
size 627
|
lora-alpaca/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,656 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.883411169052124,
|
| 3 |
+
"best_model_checkpoint": "./lora-alpaca/checkpoint-1000",
|
| 4 |
+
"epoch": 2.002002002002002,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 1.8749,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 2.119,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 3e-05,
|
| 25 |
+
"loss": 2.3009,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08,
|
| 30 |
+
"learning_rate": 4e-05,
|
| 31 |
+
"loss": 2.3636,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.1,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 2.1696,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.12,
|
| 42 |
+
"learning_rate": 6e-05,
|
| 43 |
+
"loss": 1.4449,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 7e-05,
|
| 49 |
+
"loss": 1.3746,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.16,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 1.0467,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.18,
|
| 60 |
+
"learning_rate": 9e-05,
|
| 61 |
+
"loss": 0.7711,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.2,
|
| 66 |
+
"learning_rate": 0.0001,
|
| 67 |
+
"loss": 0.6918,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.22,
|
| 72 |
+
"learning_rate": 9.92841803865426e-05,
|
| 73 |
+
"loss": 1.0876,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.24,
|
| 78 |
+
"learning_rate": 9.856836077308519e-05,
|
| 79 |
+
"loss": 0.9898,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.26,
|
| 84 |
+
"learning_rate": 9.785254115962778e-05,
|
| 85 |
+
"loss": 0.8701,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.28,
|
| 90 |
+
"learning_rate": 9.713672154617037e-05,
|
| 91 |
+
"loss": 0.7106,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.3,
|
| 96 |
+
"learning_rate": 9.642090193271297e-05,
|
| 97 |
+
"loss": 0.6589,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.32,
|
| 102 |
+
"learning_rate": 9.570508231925556e-05,
|
| 103 |
+
"loss": 1.0803,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.34,
|
| 108 |
+
"learning_rate": 9.498926270579815e-05,
|
| 109 |
+
"loss": 0.9667,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.36,
|
| 114 |
+
"learning_rate": 9.427344309234073e-05,
|
| 115 |
+
"loss": 0.8258,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.38,
|
| 120 |
+
"learning_rate": 9.355762347888332e-05,
|
| 121 |
+
"loss": 0.6852,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.4,
|
| 126 |
+
"learning_rate": 9.284180386542591e-05,
|
| 127 |
+
"loss": 0.6465,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4,
|
| 132 |
+
"eval_loss": 0.9591875076293945,
|
| 133 |
+
"eval_runtime": 134.9228,
|
| 134 |
+
"eval_samples_per_second": 14.823,
|
| 135 |
+
"eval_steps_per_second": 1.853,
|
| 136 |
+
"step": 200
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.42,
|
| 140 |
+
"learning_rate": 9.21259842519685e-05,
|
| 141 |
+
"loss": 1.0695,
|
| 142 |
+
"step": 210
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.44,
|
| 146 |
+
"learning_rate": 9.14101646385111e-05,
|
| 147 |
+
"loss": 0.9637,
|
| 148 |
+
"step": 220
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.46,
|
| 152 |
+
"learning_rate": 9.069434502505369e-05,
|
| 153 |
+
"loss": 0.8331,
|
| 154 |
+
"step": 230
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.48,
|
| 158 |
+
"learning_rate": 8.997852541159628e-05,
|
| 159 |
+
"loss": 0.6924,
|
| 160 |
+
"step": 240
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.5,
|
| 164 |
+
"learning_rate": 8.926270579813887e-05,
|
| 165 |
+
"loss": 0.6501,
|
| 166 |
+
"step": 250
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.52,
|
| 170 |
+
"learning_rate": 8.854688618468147e-05,
|
| 171 |
+
"loss": 1.0621,
|
| 172 |
+
"step": 260
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.54,
|
| 176 |
+
"learning_rate": 8.783106657122406e-05,
|
| 177 |
+
"loss": 0.9574,
|
| 178 |
+
"step": 270
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.56,
|
| 182 |
+
"learning_rate": 8.711524695776665e-05,
|
| 183 |
+
"loss": 0.8269,
|
| 184 |
+
"step": 280
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.58,
|
| 188 |
+
"learning_rate": 8.639942734430924e-05,
|
| 189 |
+
"loss": 0.6733,
|
| 190 |
+
"step": 290
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.6,
|
| 194 |
+
"learning_rate": 8.568360773085183e-05,
|
| 195 |
+
"loss": 0.6303,
|
| 196 |
+
"step": 300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.62,
|
| 200 |
+
"learning_rate": 8.496778811739443e-05,
|
| 201 |
+
"loss": 1.0696,
|
| 202 |
+
"step": 310
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.64,
|
| 206 |
+
"learning_rate": 8.4251968503937e-05,
|
| 207 |
+
"loss": 0.9427,
|
| 208 |
+
"step": 320
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.66,
|
| 212 |
+
"learning_rate": 8.35361488904796e-05,
|
| 213 |
+
"loss": 0.8269,
|
| 214 |
+
"step": 330
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.68,
|
| 218 |
+
"learning_rate": 8.282032927702219e-05,
|
| 219 |
+
"loss": 0.6791,
|
| 220 |
+
"step": 340
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.7,
|
| 224 |
+
"learning_rate": 8.210450966356478e-05,
|
| 225 |
+
"loss": 0.6343,
|
| 226 |
+
"step": 350
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.72,
|
| 230 |
+
"learning_rate": 8.138869005010737e-05,
|
| 231 |
+
"loss": 1.0558,
|
| 232 |
+
"step": 360
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.74,
|
| 236 |
+
"learning_rate": 8.067287043664997e-05,
|
| 237 |
+
"loss": 0.9428,
|
| 238 |
+
"step": 370
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.76,
|
| 242 |
+
"learning_rate": 7.995705082319256e-05,
|
| 243 |
+
"loss": 0.8074,
|
| 244 |
+
"step": 380
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.78,
|
| 248 |
+
"learning_rate": 7.924123120973515e-05,
|
| 249 |
+
"loss": 0.6895,
|
| 250 |
+
"step": 390
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.8,
|
| 254 |
+
"learning_rate": 7.852541159627774e-05,
|
| 255 |
+
"loss": 0.632,
|
| 256 |
+
"step": 400
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8,
|
| 260 |
+
"eval_loss": 0.920779287815094,
|
| 261 |
+
"eval_runtime": 134.2323,
|
| 262 |
+
"eval_samples_per_second": 14.9,
|
| 263 |
+
"eval_steps_per_second": 1.862,
|
| 264 |
+
"step": 400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.82,
|
| 268 |
+
"learning_rate": 7.780959198282033e-05,
|
| 269 |
+
"loss": 1.0575,
|
| 270 |
+
"step": 410
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.84,
|
| 274 |
+
"learning_rate": 7.709377236936293e-05,
|
| 275 |
+
"loss": 0.9369,
|
| 276 |
+
"step": 420
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.86,
|
| 280 |
+
"learning_rate": 7.637795275590552e-05,
|
| 281 |
+
"loss": 0.8319,
|
| 282 |
+
"step": 430
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.88,
|
| 286 |
+
"learning_rate": 7.566213314244811e-05,
|
| 287 |
+
"loss": 0.6667,
|
| 288 |
+
"step": 440
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.9,
|
| 292 |
+
"learning_rate": 7.49463135289907e-05,
|
| 293 |
+
"loss": 0.6284,
|
| 294 |
+
"step": 450
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.92,
|
| 298 |
+
"learning_rate": 7.42304939155333e-05,
|
| 299 |
+
"loss": 1.0348,
|
| 300 |
+
"step": 460
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.94,
|
| 304 |
+
"learning_rate": 7.351467430207589e-05,
|
| 305 |
+
"loss": 0.9361,
|
| 306 |
+
"step": 470
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.96,
|
| 310 |
+
"learning_rate": 7.279885468861847e-05,
|
| 311 |
+
"loss": 0.8108,
|
| 312 |
+
"step": 480
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.98,
|
| 316 |
+
"learning_rate": 7.208303507516106e-05,
|
| 317 |
+
"loss": 0.677,
|
| 318 |
+
"step": 490
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.0,
|
| 322 |
+
"learning_rate": 7.136721546170365e-05,
|
| 323 |
+
"loss": 0.6402,
|
| 324 |
+
"step": 500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.02,
|
| 328 |
+
"learning_rate": 7.065139584824624e-05,
|
| 329 |
+
"loss": 1.0431,
|
| 330 |
+
"step": 510
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.04,
|
| 334 |
+
"learning_rate": 6.993557623478884e-05,
|
| 335 |
+
"loss": 0.9406,
|
| 336 |
+
"step": 520
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.06,
|
| 340 |
+
"learning_rate": 6.921975662133143e-05,
|
| 341 |
+
"loss": 0.7971,
|
| 342 |
+
"step": 530
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.08,
|
| 346 |
+
"learning_rate": 6.850393700787402e-05,
|
| 347 |
+
"loss": 0.6652,
|
| 348 |
+
"step": 540
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.1,
|
| 352 |
+
"learning_rate": 6.778811739441661e-05,
|
| 353 |
+
"loss": 0.651,
|
| 354 |
+
"step": 550
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.12,
|
| 358 |
+
"learning_rate": 6.70722977809592e-05,
|
| 359 |
+
"loss": 1.0322,
|
| 360 |
+
"step": 560
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.14,
|
| 364 |
+
"learning_rate": 6.63564781675018e-05,
|
| 365 |
+
"loss": 0.9425,
|
| 366 |
+
"step": 570
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.16,
|
| 370 |
+
"learning_rate": 6.564065855404439e-05,
|
| 371 |
+
"loss": 0.8085,
|
| 372 |
+
"step": 580
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 1.18,
|
| 376 |
+
"learning_rate": 6.492483894058698e-05,
|
| 377 |
+
"loss": 0.6654,
|
| 378 |
+
"step": 590
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.2,
|
| 382 |
+
"learning_rate": 6.420901932712957e-05,
|
| 383 |
+
"loss": 0.6493,
|
| 384 |
+
"step": 600
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.2,
|
| 388 |
+
"eval_loss": 0.9059903025627136,
|
| 389 |
+
"eval_runtime": 133.8487,
|
| 390 |
+
"eval_samples_per_second": 14.942,
|
| 391 |
+
"eval_steps_per_second": 1.868,
|
| 392 |
+
"step": 600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.22,
|
| 396 |
+
"learning_rate": 6.349319971367216e-05,
|
| 397 |
+
"loss": 1.0063,
|
| 398 |
+
"step": 610
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.24,
|
| 402 |
+
"learning_rate": 6.277738010021474e-05,
|
| 403 |
+
"loss": 0.9179,
|
| 404 |
+
"step": 620
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.26,
|
| 408 |
+
"learning_rate": 6.206156048675734e-05,
|
| 409 |
+
"loss": 0.797,
|
| 410 |
+
"step": 630
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.28,
|
| 414 |
+
"learning_rate": 6.134574087329993e-05,
|
| 415 |
+
"loss": 0.6743,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.3,
|
| 420 |
+
"learning_rate": 6.0629921259842526e-05,
|
| 421 |
+
"loss": 0.6468,
|
| 422 |
+
"step": 650
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.32,
|
| 426 |
+
"learning_rate": 5.991410164638511e-05,
|
| 427 |
+
"loss": 1.0237,
|
| 428 |
+
"step": 660
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.34,
|
| 432 |
+
"learning_rate": 5.919828203292771e-05,
|
| 433 |
+
"loss": 0.9141,
|
| 434 |
+
"step": 670
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.36,
|
| 438 |
+
"learning_rate": 5.8482462419470296e-05,
|
| 439 |
+
"loss": 0.7917,
|
| 440 |
+
"step": 680
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.38,
|
| 444 |
+
"learning_rate": 5.776664280601288e-05,
|
| 445 |
+
"loss": 0.6499,
|
| 446 |
+
"step": 690
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.4,
|
| 450 |
+
"learning_rate": 5.705082319255548e-05,
|
| 451 |
+
"loss": 0.6481,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.42,
|
| 456 |
+
"learning_rate": 5.6335003579098066e-05,
|
| 457 |
+
"loss": 1.0164,
|
| 458 |
+
"step": 710
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.44,
|
| 462 |
+
"learning_rate": 5.5619183965640665e-05,
|
| 463 |
+
"loss": 0.9252,
|
| 464 |
+
"step": 720
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.46,
|
| 468 |
+
"learning_rate": 5.490336435218325e-05,
|
| 469 |
+
"loss": 0.7942,
|
| 470 |
+
"step": 730
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.48,
|
| 474 |
+
"learning_rate": 5.418754473872585e-05,
|
| 475 |
+
"loss": 0.6517,
|
| 476 |
+
"step": 740
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.5,
|
| 480 |
+
"learning_rate": 5.3471725125268434e-05,
|
| 481 |
+
"loss": 0.642,
|
| 482 |
+
"step": 750
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.52,
|
| 486 |
+
"learning_rate": 5.275590551181102e-05,
|
| 487 |
+
"loss": 1.0249,
|
| 488 |
+
"step": 760
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.54,
|
| 492 |
+
"learning_rate": 5.204008589835362e-05,
|
| 493 |
+
"loss": 0.9345,
|
| 494 |
+
"step": 770
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.56,
|
| 498 |
+
"learning_rate": 5.1324266284896204e-05,
|
| 499 |
+
"loss": 0.7841,
|
| 500 |
+
"step": 780
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.58,
|
| 504 |
+
"learning_rate": 5.06084466714388e-05,
|
| 505 |
+
"loss": 0.6512,
|
| 506 |
+
"step": 790
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6,
|
| 510 |
+
"learning_rate": 4.989262705798139e-05,
|
| 511 |
+
"loss": 0.6472,
|
| 512 |
+
"step": 800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.6,
|
| 516 |
+
"eval_loss": 0.8923133611679077,
|
| 517 |
+
"eval_runtime": 133.1272,
|
| 518 |
+
"eval_samples_per_second": 15.023,
|
| 519 |
+
"eval_steps_per_second": 1.878,
|
| 520 |
+
"step": 800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.62,
|
| 524 |
+
"learning_rate": 4.917680744452398e-05,
|
| 525 |
+
"loss": 1.0134,
|
| 526 |
+
"step": 810
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.64,
|
| 530 |
+
"learning_rate": 4.846098783106657e-05,
|
| 531 |
+
"loss": 0.9177,
|
| 532 |
+
"step": 820
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.66,
|
| 536 |
+
"learning_rate": 4.7745168217609165e-05,
|
| 537 |
+
"loss": 0.786,
|
| 538 |
+
"step": 830
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.68,
|
| 542 |
+
"learning_rate": 4.702934860415176e-05,
|
| 543 |
+
"loss": 0.6569,
|
| 544 |
+
"step": 840
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.7,
|
| 548 |
+
"learning_rate": 4.631352899069435e-05,
|
| 549 |
+
"loss": 0.6258,
|
| 550 |
+
"step": 850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.72,
|
| 554 |
+
"learning_rate": 4.559770937723694e-05,
|
| 555 |
+
"loss": 1.0336,
|
| 556 |
+
"step": 860
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.74,
|
| 560 |
+
"learning_rate": 4.488188976377953e-05,
|
| 561 |
+
"loss": 0.9346,
|
| 562 |
+
"step": 870
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.76,
|
| 566 |
+
"learning_rate": 4.416607015032212e-05,
|
| 567 |
+
"loss": 0.7893,
|
| 568 |
+
"step": 880
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.78,
|
| 572 |
+
"learning_rate": 4.345025053686471e-05,
|
| 573 |
+
"loss": 0.6585,
|
| 574 |
+
"step": 890
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.8,
|
| 578 |
+
"learning_rate": 4.27344309234073e-05,
|
| 579 |
+
"loss": 0.6222,
|
| 580 |
+
"step": 900
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.82,
|
| 584 |
+
"learning_rate": 4.2018611309949896e-05,
|
| 585 |
+
"loss": 1.0196,
|
| 586 |
+
"step": 910
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.84,
|
| 590 |
+
"learning_rate": 4.130279169649249e-05,
|
| 591 |
+
"loss": 0.9398,
|
| 592 |
+
"step": 920
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.86,
|
| 596 |
+
"learning_rate": 4.058697208303508e-05,
|
| 597 |
+
"loss": 0.8012,
|
| 598 |
+
"step": 930
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.88,
|
| 602 |
+
"learning_rate": 3.9871152469577665e-05,
|
| 603 |
+
"loss": 0.6599,
|
| 604 |
+
"step": 940
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.9,
|
| 608 |
+
"learning_rate": 3.915533285612026e-05,
|
| 609 |
+
"loss": 0.6337,
|
| 610 |
+
"step": 950
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.92,
|
| 614 |
+
"learning_rate": 3.843951324266285e-05,
|
| 615 |
+
"loss": 1.0181,
|
| 616 |
+
"step": 960
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.94,
|
| 620 |
+
"learning_rate": 3.772369362920544e-05,
|
| 621 |
+
"loss": 0.9119,
|
| 622 |
+
"step": 970
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.96,
|
| 626 |
+
"learning_rate": 3.7007874015748034e-05,
|
| 627 |
+
"loss": 0.7974,
|
| 628 |
+
"step": 980
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.98,
|
| 632 |
+
"learning_rate": 3.6292054402290626e-05,
|
| 633 |
+
"loss": 0.6506,
|
| 634 |
+
"step": 990
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.0,
|
| 638 |
+
"learning_rate": 3.557623478883322e-05,
|
| 639 |
+
"loss": 0.6552,
|
| 640 |
+
"step": 1000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.0,
|
| 644 |
+
"eval_loss": 0.883411169052124,
|
| 645 |
+
"eval_runtime": 133.1959,
|
| 646 |
+
"eval_samples_per_second": 15.015,
|
| 647 |
+
"eval_steps_per_second": 1.877,
|
| 648 |
+
"step": 1000
|
| 649 |
+
}
|
| 650 |
+
],
|
| 651 |
+
"max_steps": 1497,
|
| 652 |
+
"num_train_epochs": 3,
|
| 653 |
+
"total_flos": 5.306117790616781e+17,
|
| 654 |
+
"trial_name": null,
|
| 655 |
+
"trial_params": null
|
| 656 |
+
}
|
lora-alpaca/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3365a9fddc28ac19526733cfa419ff1af8b6a9569c78e529ea4a0e67de3e3a30
|
| 3 |
+
size 3515
|
lora-alpaca/checkpoint-1200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f92e82496257c26bb561412b18a0481f420f58654d2438a634e3ab28bda33b6c
|
| 3 |
+
size 33629893
|
lora-alpaca/checkpoint-1200/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e10e5f41ed20a15af2ca9b09b05e39da32caf6a28becdbd6ed3b074d8c74f3a
|
| 3 |
+
size 16822989
|
lora-alpaca/checkpoint-1200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e06bc74c1412f57258e7ccbb8ab09fb4f704e1f8c4acb61fd0a6e6c7c19ff2e
|
| 3 |
+
size 14575
|
lora-alpaca/checkpoint-1200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741cefeca9ef427f92406d2d10b81996655e2a9d50eb7aaa9614e6fdd1c9f529
|
| 3 |
+
size 557
|
lora-alpaca/checkpoint-1200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ada432692da53a0d5a6295ff4ad7a465bf6376ad1f953e6a8aa133fb848aad44
|
| 3 |
+
size 627
|
lora-alpaca/checkpoint-1200/trainer_state.json
ADDED
|
@@ -0,0 +1,784 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8776447176933289,
|
| 3 |
+
"best_model_checkpoint": "./lora-alpaca/checkpoint-1200",
|
| 4 |
+
"epoch": 2.4024024024024024,
|
| 5 |
+
"global_step": 1200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 1.8749,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 2.119,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 3e-05,
|
| 25 |
+
"loss": 2.3009,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08,
|
| 30 |
+
"learning_rate": 4e-05,
|
| 31 |
+
"loss": 2.3636,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.1,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 2.1696,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.12,
|
| 42 |
+
"learning_rate": 6e-05,
|
| 43 |
+
"loss": 1.4449,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 7e-05,
|
| 49 |
+
"loss": 1.3746,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.16,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 1.0467,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.18,
|
| 60 |
+
"learning_rate": 9e-05,
|
| 61 |
+
"loss": 0.7711,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.2,
|
| 66 |
+
"learning_rate": 0.0001,
|
| 67 |
+
"loss": 0.6918,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.22,
|
| 72 |
+
"learning_rate": 9.92841803865426e-05,
|
| 73 |
+
"loss": 1.0876,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.24,
|
| 78 |
+
"learning_rate": 9.856836077308519e-05,
|
| 79 |
+
"loss": 0.9898,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.26,
|
| 84 |
+
"learning_rate": 9.785254115962778e-05,
|
| 85 |
+
"loss": 0.8701,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.28,
|
| 90 |
+
"learning_rate": 9.713672154617037e-05,
|
| 91 |
+
"loss": 0.7106,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.3,
|
| 96 |
+
"learning_rate": 9.642090193271297e-05,
|
| 97 |
+
"loss": 0.6589,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.32,
|
| 102 |
+
"learning_rate": 9.570508231925556e-05,
|
| 103 |
+
"loss": 1.0803,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.34,
|
| 108 |
+
"learning_rate": 9.498926270579815e-05,
|
| 109 |
+
"loss": 0.9667,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.36,
|
| 114 |
+
"learning_rate": 9.427344309234073e-05,
|
| 115 |
+
"loss": 0.8258,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.38,
|
| 120 |
+
"learning_rate": 9.355762347888332e-05,
|
| 121 |
+
"loss": 0.6852,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.4,
|
| 126 |
+
"learning_rate": 9.284180386542591e-05,
|
| 127 |
+
"loss": 0.6465,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4,
|
| 132 |
+
"eval_loss": 0.9591875076293945,
|
| 133 |
+
"eval_runtime": 134.9228,
|
| 134 |
+
"eval_samples_per_second": 14.823,
|
| 135 |
+
"eval_steps_per_second": 1.853,
|
| 136 |
+
"step": 200
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.42,
|
| 140 |
+
"learning_rate": 9.21259842519685e-05,
|
| 141 |
+
"loss": 1.0695,
|
| 142 |
+
"step": 210
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.44,
|
| 146 |
+
"learning_rate": 9.14101646385111e-05,
|
| 147 |
+
"loss": 0.9637,
|
| 148 |
+
"step": 220
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.46,
|
| 152 |
+
"learning_rate": 9.069434502505369e-05,
|
| 153 |
+
"loss": 0.8331,
|
| 154 |
+
"step": 230
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.48,
|
| 158 |
+
"learning_rate": 8.997852541159628e-05,
|
| 159 |
+
"loss": 0.6924,
|
| 160 |
+
"step": 240
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.5,
|
| 164 |
+
"learning_rate": 8.926270579813887e-05,
|
| 165 |
+
"loss": 0.6501,
|
| 166 |
+
"step": 250
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.52,
|
| 170 |
+
"learning_rate": 8.854688618468147e-05,
|
| 171 |
+
"loss": 1.0621,
|
| 172 |
+
"step": 260
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.54,
|
| 176 |
+
"learning_rate": 8.783106657122406e-05,
|
| 177 |
+
"loss": 0.9574,
|
| 178 |
+
"step": 270
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.56,
|
| 182 |
+
"learning_rate": 8.711524695776665e-05,
|
| 183 |
+
"loss": 0.8269,
|
| 184 |
+
"step": 280
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.58,
|
| 188 |
+
"learning_rate": 8.639942734430924e-05,
|
| 189 |
+
"loss": 0.6733,
|
| 190 |
+
"step": 290
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.6,
|
| 194 |
+
"learning_rate": 8.568360773085183e-05,
|
| 195 |
+
"loss": 0.6303,
|
| 196 |
+
"step": 300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.62,
|
| 200 |
+
"learning_rate": 8.496778811739443e-05,
|
| 201 |
+
"loss": 1.0696,
|
| 202 |
+
"step": 310
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.64,
|
| 206 |
+
"learning_rate": 8.4251968503937e-05,
|
| 207 |
+
"loss": 0.9427,
|
| 208 |
+
"step": 320
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.66,
|
| 212 |
+
"learning_rate": 8.35361488904796e-05,
|
| 213 |
+
"loss": 0.8269,
|
| 214 |
+
"step": 330
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.68,
|
| 218 |
+
"learning_rate": 8.282032927702219e-05,
|
| 219 |
+
"loss": 0.6791,
|
| 220 |
+
"step": 340
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.7,
|
| 224 |
+
"learning_rate": 8.210450966356478e-05,
|
| 225 |
+
"loss": 0.6343,
|
| 226 |
+
"step": 350
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.72,
|
| 230 |
+
"learning_rate": 8.138869005010737e-05,
|
| 231 |
+
"loss": 1.0558,
|
| 232 |
+
"step": 360
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.74,
|
| 236 |
+
"learning_rate": 8.067287043664997e-05,
|
| 237 |
+
"loss": 0.9428,
|
| 238 |
+
"step": 370
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.76,
|
| 242 |
+
"learning_rate": 7.995705082319256e-05,
|
| 243 |
+
"loss": 0.8074,
|
| 244 |
+
"step": 380
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.78,
|
| 248 |
+
"learning_rate": 7.924123120973515e-05,
|
| 249 |
+
"loss": 0.6895,
|
| 250 |
+
"step": 390
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.8,
|
| 254 |
+
"learning_rate": 7.852541159627774e-05,
|
| 255 |
+
"loss": 0.632,
|
| 256 |
+
"step": 400
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8,
|
| 260 |
+
"eval_loss": 0.920779287815094,
|
| 261 |
+
"eval_runtime": 134.2323,
|
| 262 |
+
"eval_samples_per_second": 14.9,
|
| 263 |
+
"eval_steps_per_second": 1.862,
|
| 264 |
+
"step": 400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.82,
|
| 268 |
+
"learning_rate": 7.780959198282033e-05,
|
| 269 |
+
"loss": 1.0575,
|
| 270 |
+
"step": 410
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.84,
|
| 274 |
+
"learning_rate": 7.709377236936293e-05,
|
| 275 |
+
"loss": 0.9369,
|
| 276 |
+
"step": 420
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.86,
|
| 280 |
+
"learning_rate": 7.637795275590552e-05,
|
| 281 |
+
"loss": 0.8319,
|
| 282 |
+
"step": 430
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.88,
|
| 286 |
+
"learning_rate": 7.566213314244811e-05,
|
| 287 |
+
"loss": 0.6667,
|
| 288 |
+
"step": 440
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.9,
|
| 292 |
+
"learning_rate": 7.49463135289907e-05,
|
| 293 |
+
"loss": 0.6284,
|
| 294 |
+
"step": 450
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.92,
|
| 298 |
+
"learning_rate": 7.42304939155333e-05,
|
| 299 |
+
"loss": 1.0348,
|
| 300 |
+
"step": 460
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.94,
|
| 304 |
+
"learning_rate": 7.351467430207589e-05,
|
| 305 |
+
"loss": 0.9361,
|
| 306 |
+
"step": 470
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.96,
|
| 310 |
+
"learning_rate": 7.279885468861847e-05,
|
| 311 |
+
"loss": 0.8108,
|
| 312 |
+
"step": 480
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.98,
|
| 316 |
+
"learning_rate": 7.208303507516106e-05,
|
| 317 |
+
"loss": 0.677,
|
| 318 |
+
"step": 490
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.0,
|
| 322 |
+
"learning_rate": 7.136721546170365e-05,
|
| 323 |
+
"loss": 0.6402,
|
| 324 |
+
"step": 500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.02,
|
| 328 |
+
"learning_rate": 7.065139584824624e-05,
|
| 329 |
+
"loss": 1.0431,
|
| 330 |
+
"step": 510
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.04,
|
| 334 |
+
"learning_rate": 6.993557623478884e-05,
|
| 335 |
+
"loss": 0.9406,
|
| 336 |
+
"step": 520
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.06,
|
| 340 |
+
"learning_rate": 6.921975662133143e-05,
|
| 341 |
+
"loss": 0.7971,
|
| 342 |
+
"step": 530
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.08,
|
| 346 |
+
"learning_rate": 6.850393700787402e-05,
|
| 347 |
+
"loss": 0.6652,
|
| 348 |
+
"step": 540
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.1,
|
| 352 |
+
"learning_rate": 6.778811739441661e-05,
|
| 353 |
+
"loss": 0.651,
|
| 354 |
+
"step": 550
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.12,
|
| 358 |
+
"learning_rate": 6.70722977809592e-05,
|
| 359 |
+
"loss": 1.0322,
|
| 360 |
+
"step": 560
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.14,
|
| 364 |
+
"learning_rate": 6.63564781675018e-05,
|
| 365 |
+
"loss": 0.9425,
|
| 366 |
+
"step": 570
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.16,
|
| 370 |
+
"learning_rate": 6.564065855404439e-05,
|
| 371 |
+
"loss": 0.8085,
|
| 372 |
+
"step": 580
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 1.18,
|
| 376 |
+
"learning_rate": 6.492483894058698e-05,
|
| 377 |
+
"loss": 0.6654,
|
| 378 |
+
"step": 590
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.2,
|
| 382 |
+
"learning_rate": 6.420901932712957e-05,
|
| 383 |
+
"loss": 0.6493,
|
| 384 |
+
"step": 600
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.2,
|
| 388 |
+
"eval_loss": 0.9059903025627136,
|
| 389 |
+
"eval_runtime": 133.8487,
|
| 390 |
+
"eval_samples_per_second": 14.942,
|
| 391 |
+
"eval_steps_per_second": 1.868,
|
| 392 |
+
"step": 600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.22,
|
| 396 |
+
"learning_rate": 6.349319971367216e-05,
|
| 397 |
+
"loss": 1.0063,
|
| 398 |
+
"step": 610
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.24,
|
| 402 |
+
"learning_rate": 6.277738010021474e-05,
|
| 403 |
+
"loss": 0.9179,
|
| 404 |
+
"step": 620
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.26,
|
| 408 |
+
"learning_rate": 6.206156048675734e-05,
|
| 409 |
+
"loss": 0.797,
|
| 410 |
+
"step": 630
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.28,
|
| 414 |
+
"learning_rate": 6.134574087329993e-05,
|
| 415 |
+
"loss": 0.6743,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.3,
|
| 420 |
+
"learning_rate": 6.0629921259842526e-05,
|
| 421 |
+
"loss": 0.6468,
|
| 422 |
+
"step": 650
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.32,
|
| 426 |
+
"learning_rate": 5.991410164638511e-05,
|
| 427 |
+
"loss": 1.0237,
|
| 428 |
+
"step": 660
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.34,
|
| 432 |
+
"learning_rate": 5.919828203292771e-05,
|
| 433 |
+
"loss": 0.9141,
|
| 434 |
+
"step": 670
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.36,
|
| 438 |
+
"learning_rate": 5.8482462419470296e-05,
|
| 439 |
+
"loss": 0.7917,
|
| 440 |
+
"step": 680
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.38,
|
| 444 |
+
"learning_rate": 5.776664280601288e-05,
|
| 445 |
+
"loss": 0.6499,
|
| 446 |
+
"step": 690
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.4,
|
| 450 |
+
"learning_rate": 5.705082319255548e-05,
|
| 451 |
+
"loss": 0.6481,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.42,
|
| 456 |
+
"learning_rate": 5.6335003579098066e-05,
|
| 457 |
+
"loss": 1.0164,
|
| 458 |
+
"step": 710
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.44,
|
| 462 |
+
"learning_rate": 5.5619183965640665e-05,
|
| 463 |
+
"loss": 0.9252,
|
| 464 |
+
"step": 720
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.46,
|
| 468 |
+
"learning_rate": 5.490336435218325e-05,
|
| 469 |
+
"loss": 0.7942,
|
| 470 |
+
"step": 730
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.48,
|
| 474 |
+
"learning_rate": 5.418754473872585e-05,
|
| 475 |
+
"loss": 0.6517,
|
| 476 |
+
"step": 740
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.5,
|
| 480 |
+
"learning_rate": 5.3471725125268434e-05,
|
| 481 |
+
"loss": 0.642,
|
| 482 |
+
"step": 750
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.52,
|
| 486 |
+
"learning_rate": 5.275590551181102e-05,
|
| 487 |
+
"loss": 1.0249,
|
| 488 |
+
"step": 760
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.54,
|
| 492 |
+
"learning_rate": 5.204008589835362e-05,
|
| 493 |
+
"loss": 0.9345,
|
| 494 |
+
"step": 770
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.56,
|
| 498 |
+
"learning_rate": 5.1324266284896204e-05,
|
| 499 |
+
"loss": 0.7841,
|
| 500 |
+
"step": 780
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.58,
|
| 504 |
+
"learning_rate": 5.06084466714388e-05,
|
| 505 |
+
"loss": 0.6512,
|
| 506 |
+
"step": 790
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6,
|
| 510 |
+
"learning_rate": 4.989262705798139e-05,
|
| 511 |
+
"loss": 0.6472,
|
| 512 |
+
"step": 800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.6,
|
| 516 |
+
"eval_loss": 0.8923133611679077,
|
| 517 |
+
"eval_runtime": 133.1272,
|
| 518 |
+
"eval_samples_per_second": 15.023,
|
| 519 |
+
"eval_steps_per_second": 1.878,
|
| 520 |
+
"step": 800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.62,
|
| 524 |
+
"learning_rate": 4.917680744452398e-05,
|
| 525 |
+
"loss": 1.0134,
|
| 526 |
+
"step": 810
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.64,
|
| 530 |
+
"learning_rate": 4.846098783106657e-05,
|
| 531 |
+
"loss": 0.9177,
|
| 532 |
+
"step": 820
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.66,
|
| 536 |
+
"learning_rate": 4.7745168217609165e-05,
|
| 537 |
+
"loss": 0.786,
|
| 538 |
+
"step": 830
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.68,
|
| 542 |
+
"learning_rate": 4.702934860415176e-05,
|
| 543 |
+
"loss": 0.6569,
|
| 544 |
+
"step": 840
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.7,
|
| 548 |
+
"learning_rate": 4.631352899069435e-05,
|
| 549 |
+
"loss": 0.6258,
|
| 550 |
+
"step": 850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.72,
|
| 554 |
+
"learning_rate": 4.559770937723694e-05,
|
| 555 |
+
"loss": 1.0336,
|
| 556 |
+
"step": 860
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.74,
|
| 560 |
+
"learning_rate": 4.488188976377953e-05,
|
| 561 |
+
"loss": 0.9346,
|
| 562 |
+
"step": 870
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.76,
|
| 566 |
+
"learning_rate": 4.416607015032212e-05,
|
| 567 |
+
"loss": 0.7893,
|
| 568 |
+
"step": 880
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.78,
|
| 572 |
+
"learning_rate": 4.345025053686471e-05,
|
| 573 |
+
"loss": 0.6585,
|
| 574 |
+
"step": 890
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.8,
|
| 578 |
+
"learning_rate": 4.27344309234073e-05,
|
| 579 |
+
"loss": 0.6222,
|
| 580 |
+
"step": 900
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.82,
|
| 584 |
+
"learning_rate": 4.2018611309949896e-05,
|
| 585 |
+
"loss": 1.0196,
|
| 586 |
+
"step": 910
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.84,
|
| 590 |
+
"learning_rate": 4.130279169649249e-05,
|
| 591 |
+
"loss": 0.9398,
|
| 592 |
+
"step": 920
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.86,
|
| 596 |
+
"learning_rate": 4.058697208303508e-05,
|
| 597 |
+
"loss": 0.8012,
|
| 598 |
+
"step": 930
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.88,
|
| 602 |
+
"learning_rate": 3.9871152469577665e-05,
|
| 603 |
+
"loss": 0.6599,
|
| 604 |
+
"step": 940
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.9,
|
| 608 |
+
"learning_rate": 3.915533285612026e-05,
|
| 609 |
+
"loss": 0.6337,
|
| 610 |
+
"step": 950
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.92,
|
| 614 |
+
"learning_rate": 3.843951324266285e-05,
|
| 615 |
+
"loss": 1.0181,
|
| 616 |
+
"step": 960
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.94,
|
| 620 |
+
"learning_rate": 3.772369362920544e-05,
|
| 621 |
+
"loss": 0.9119,
|
| 622 |
+
"step": 970
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.96,
|
| 626 |
+
"learning_rate": 3.7007874015748034e-05,
|
| 627 |
+
"loss": 0.7974,
|
| 628 |
+
"step": 980
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.98,
|
| 632 |
+
"learning_rate": 3.6292054402290626e-05,
|
| 633 |
+
"loss": 0.6506,
|
| 634 |
+
"step": 990
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.0,
|
| 638 |
+
"learning_rate": 3.557623478883322e-05,
|
| 639 |
+
"loss": 0.6552,
|
| 640 |
+
"step": 1000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.0,
|
| 644 |
+
"eval_loss": 0.883411169052124,
|
| 645 |
+
"eval_runtime": 133.1959,
|
| 646 |
+
"eval_samples_per_second": 15.015,
|
| 647 |
+
"eval_steps_per_second": 1.877,
|
| 648 |
+
"step": 1000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.02,
|
| 652 |
+
"learning_rate": 3.4860415175375804e-05,
|
| 653 |
+
"loss": 1.0039,
|
| 654 |
+
"step": 1010
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 2.04,
|
| 658 |
+
"learning_rate": 3.4144595561918396e-05,
|
| 659 |
+
"loss": 0.9259,
|
| 660 |
+
"step": 1020
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.06,
|
| 664 |
+
"learning_rate": 3.342877594846099e-05,
|
| 665 |
+
"loss": 0.7858,
|
| 666 |
+
"step": 1030
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 2.08,
|
| 670 |
+
"learning_rate": 3.271295633500358e-05,
|
| 671 |
+
"loss": 0.649,
|
| 672 |
+
"step": 1040
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.1,
|
| 676 |
+
"learning_rate": 3.199713672154617e-05,
|
| 677 |
+
"loss": 0.6557,
|
| 678 |
+
"step": 1050
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.12,
|
| 682 |
+
"learning_rate": 3.1281317108088764e-05,
|
| 683 |
+
"loss": 1.0129,
|
| 684 |
+
"step": 1060
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.14,
|
| 688 |
+
"learning_rate": 3.0565497494631357e-05,
|
| 689 |
+
"loss": 0.901,
|
| 690 |
+
"step": 1070
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.16,
|
| 694 |
+
"learning_rate": 2.9849677881173942e-05,
|
| 695 |
+
"loss": 0.7697,
|
| 696 |
+
"step": 1080
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 2.18,
|
| 700 |
+
"learning_rate": 2.9133858267716534e-05,
|
| 701 |
+
"loss": 0.6419,
|
| 702 |
+
"step": 1090
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.2,
|
| 706 |
+
"learning_rate": 2.8418038654259126e-05,
|
| 707 |
+
"loss": 0.6546,
|
| 708 |
+
"step": 1100
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 2.22,
|
| 712 |
+
"learning_rate": 2.770221904080172e-05,
|
| 713 |
+
"loss": 1.009,
|
| 714 |
+
"step": 1110
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.24,
|
| 718 |
+
"learning_rate": 2.698639942734431e-05,
|
| 719 |
+
"loss": 0.9193,
|
| 720 |
+
"step": 1120
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.26,
|
| 724 |
+
"learning_rate": 2.6270579813886903e-05,
|
| 725 |
+
"loss": 0.7808,
|
| 726 |
+
"step": 1130
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.28,
|
| 730 |
+
"learning_rate": 2.5554760200429495e-05,
|
| 731 |
+
"loss": 0.6449,
|
| 732 |
+
"step": 1140
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.3,
|
| 736 |
+
"learning_rate": 2.4838940586972084e-05,
|
| 737 |
+
"loss": 0.6551,
|
| 738 |
+
"step": 1150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.32,
|
| 742 |
+
"learning_rate": 2.4123120973514676e-05,
|
| 743 |
+
"loss": 1.0013,
|
| 744 |
+
"step": 1160
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.34,
|
| 748 |
+
"learning_rate": 2.3407301360057265e-05,
|
| 749 |
+
"loss": 0.9067,
|
| 750 |
+
"step": 1170
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.36,
|
| 754 |
+
"learning_rate": 2.2691481746599857e-05,
|
| 755 |
+
"loss": 0.7739,
|
| 756 |
+
"step": 1180
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 2.38,
|
| 760 |
+
"learning_rate": 2.197566213314245e-05,
|
| 761 |
+
"loss": 0.6502,
|
| 762 |
+
"step": 1190
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.4,
|
| 766 |
+
"learning_rate": 2.125984251968504e-05,
|
| 767 |
+
"loss": 0.6557,
|
| 768 |
+
"step": 1200
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.4,
|
| 772 |
+
"eval_loss": 0.8776447176933289,
|
| 773 |
+
"eval_runtime": 133.098,
|
| 774 |
+
"eval_samples_per_second": 15.027,
|
| 775 |
+
"eval_steps_per_second": 1.878,
|
| 776 |
+
"step": 1200
|
| 777 |
+
}
|
| 778 |
+
],
|
| 779 |
+
"max_steps": 1497,
|
| 780 |
+
"num_train_epochs": 3,
|
| 781 |
+
"total_flos": 6.359415119727821e+17,
|
| 782 |
+
"trial_name": null,
|
| 783 |
+
"trial_params": null
|
| 784 |
+
}
|
lora-alpaca/checkpoint-1200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3365a9fddc28ac19526733cfa419ff1af8b6a9569c78e529ea4a0e67de3e3a30
|
| 3 |
+
size 3515
|
lora-alpaca/checkpoint-1400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95cc267ba8f77bca3b47bc121609daf2d5c257137410bee9a4d2ca3f18fa9f66
|
| 3 |
+
size 33629893
|
lora-alpaca/checkpoint-1400/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98987187f0889b1d424836d31b54105325ebb265401b6f7443327554073c3a50
|
| 3 |
+
size 16822989
|
lora-alpaca/checkpoint-1400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc5aefd83e95cf8bfeb2c064cb56777a4c09f097aab7178b02787e4e99eb4260
|
| 3 |
+
size 14575
|
lora-alpaca/checkpoint-1400/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:16fdfc03b58220402968eacaac23fb5471cdb9061302380bd3c8d4d326c02ade
|
| 3 |
+
size 557
|
lora-alpaca/checkpoint-1400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a674214d8e2d9ae93e71741636ee2e62f5b2df49177833ea6b06850296b0a93
|
| 3 |
+
size 627
|
lora-alpaca/checkpoint-1400/trainer_state.json
ADDED
|
@@ -0,0 +1,912 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8746508359909058,
|
| 3 |
+
"best_model_checkpoint": "./lora-alpaca/checkpoint-1400",
|
| 4 |
+
"epoch": 2.8028028028028027,
|
| 5 |
+
"global_step": 1400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 1.8749,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 2.119,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 3e-05,
|
| 25 |
+
"loss": 2.3009,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08,
|
| 30 |
+
"learning_rate": 4e-05,
|
| 31 |
+
"loss": 2.3636,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.1,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 2.1696,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.12,
|
| 42 |
+
"learning_rate": 6e-05,
|
| 43 |
+
"loss": 1.4449,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 7e-05,
|
| 49 |
+
"loss": 1.3746,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.16,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 1.0467,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.18,
|
| 60 |
+
"learning_rate": 9e-05,
|
| 61 |
+
"loss": 0.7711,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.2,
|
| 66 |
+
"learning_rate": 0.0001,
|
| 67 |
+
"loss": 0.6918,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.22,
|
| 72 |
+
"learning_rate": 9.92841803865426e-05,
|
| 73 |
+
"loss": 1.0876,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.24,
|
| 78 |
+
"learning_rate": 9.856836077308519e-05,
|
| 79 |
+
"loss": 0.9898,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.26,
|
| 84 |
+
"learning_rate": 9.785254115962778e-05,
|
| 85 |
+
"loss": 0.8701,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.28,
|
| 90 |
+
"learning_rate": 9.713672154617037e-05,
|
| 91 |
+
"loss": 0.7106,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.3,
|
| 96 |
+
"learning_rate": 9.642090193271297e-05,
|
| 97 |
+
"loss": 0.6589,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.32,
|
| 102 |
+
"learning_rate": 9.570508231925556e-05,
|
| 103 |
+
"loss": 1.0803,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.34,
|
| 108 |
+
"learning_rate": 9.498926270579815e-05,
|
| 109 |
+
"loss": 0.9667,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.36,
|
| 114 |
+
"learning_rate": 9.427344309234073e-05,
|
| 115 |
+
"loss": 0.8258,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.38,
|
| 120 |
+
"learning_rate": 9.355762347888332e-05,
|
| 121 |
+
"loss": 0.6852,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.4,
|
| 126 |
+
"learning_rate": 9.284180386542591e-05,
|
| 127 |
+
"loss": 0.6465,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4,
|
| 132 |
+
"eval_loss": 0.9591875076293945,
|
| 133 |
+
"eval_runtime": 134.9228,
|
| 134 |
+
"eval_samples_per_second": 14.823,
|
| 135 |
+
"eval_steps_per_second": 1.853,
|
| 136 |
+
"step": 200
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.42,
|
| 140 |
+
"learning_rate": 9.21259842519685e-05,
|
| 141 |
+
"loss": 1.0695,
|
| 142 |
+
"step": 210
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.44,
|
| 146 |
+
"learning_rate": 9.14101646385111e-05,
|
| 147 |
+
"loss": 0.9637,
|
| 148 |
+
"step": 220
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.46,
|
| 152 |
+
"learning_rate": 9.069434502505369e-05,
|
| 153 |
+
"loss": 0.8331,
|
| 154 |
+
"step": 230
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.48,
|
| 158 |
+
"learning_rate": 8.997852541159628e-05,
|
| 159 |
+
"loss": 0.6924,
|
| 160 |
+
"step": 240
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.5,
|
| 164 |
+
"learning_rate": 8.926270579813887e-05,
|
| 165 |
+
"loss": 0.6501,
|
| 166 |
+
"step": 250
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.52,
|
| 170 |
+
"learning_rate": 8.854688618468147e-05,
|
| 171 |
+
"loss": 1.0621,
|
| 172 |
+
"step": 260
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.54,
|
| 176 |
+
"learning_rate": 8.783106657122406e-05,
|
| 177 |
+
"loss": 0.9574,
|
| 178 |
+
"step": 270
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.56,
|
| 182 |
+
"learning_rate": 8.711524695776665e-05,
|
| 183 |
+
"loss": 0.8269,
|
| 184 |
+
"step": 280
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.58,
|
| 188 |
+
"learning_rate": 8.639942734430924e-05,
|
| 189 |
+
"loss": 0.6733,
|
| 190 |
+
"step": 290
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.6,
|
| 194 |
+
"learning_rate": 8.568360773085183e-05,
|
| 195 |
+
"loss": 0.6303,
|
| 196 |
+
"step": 300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.62,
|
| 200 |
+
"learning_rate": 8.496778811739443e-05,
|
| 201 |
+
"loss": 1.0696,
|
| 202 |
+
"step": 310
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.64,
|
| 206 |
+
"learning_rate": 8.4251968503937e-05,
|
| 207 |
+
"loss": 0.9427,
|
| 208 |
+
"step": 320
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.66,
|
| 212 |
+
"learning_rate": 8.35361488904796e-05,
|
| 213 |
+
"loss": 0.8269,
|
| 214 |
+
"step": 330
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.68,
|
| 218 |
+
"learning_rate": 8.282032927702219e-05,
|
| 219 |
+
"loss": 0.6791,
|
| 220 |
+
"step": 340
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.7,
|
| 224 |
+
"learning_rate": 8.210450966356478e-05,
|
| 225 |
+
"loss": 0.6343,
|
| 226 |
+
"step": 350
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.72,
|
| 230 |
+
"learning_rate": 8.138869005010737e-05,
|
| 231 |
+
"loss": 1.0558,
|
| 232 |
+
"step": 360
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.74,
|
| 236 |
+
"learning_rate": 8.067287043664997e-05,
|
| 237 |
+
"loss": 0.9428,
|
| 238 |
+
"step": 370
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.76,
|
| 242 |
+
"learning_rate": 7.995705082319256e-05,
|
| 243 |
+
"loss": 0.8074,
|
| 244 |
+
"step": 380
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.78,
|
| 248 |
+
"learning_rate": 7.924123120973515e-05,
|
| 249 |
+
"loss": 0.6895,
|
| 250 |
+
"step": 390
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.8,
|
| 254 |
+
"learning_rate": 7.852541159627774e-05,
|
| 255 |
+
"loss": 0.632,
|
| 256 |
+
"step": 400
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8,
|
| 260 |
+
"eval_loss": 0.920779287815094,
|
| 261 |
+
"eval_runtime": 134.2323,
|
| 262 |
+
"eval_samples_per_second": 14.9,
|
| 263 |
+
"eval_steps_per_second": 1.862,
|
| 264 |
+
"step": 400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.82,
|
| 268 |
+
"learning_rate": 7.780959198282033e-05,
|
| 269 |
+
"loss": 1.0575,
|
| 270 |
+
"step": 410
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.84,
|
| 274 |
+
"learning_rate": 7.709377236936293e-05,
|
| 275 |
+
"loss": 0.9369,
|
| 276 |
+
"step": 420
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.86,
|
| 280 |
+
"learning_rate": 7.637795275590552e-05,
|
| 281 |
+
"loss": 0.8319,
|
| 282 |
+
"step": 430
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.88,
|
| 286 |
+
"learning_rate": 7.566213314244811e-05,
|
| 287 |
+
"loss": 0.6667,
|
| 288 |
+
"step": 440
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.9,
|
| 292 |
+
"learning_rate": 7.49463135289907e-05,
|
| 293 |
+
"loss": 0.6284,
|
| 294 |
+
"step": 450
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.92,
|
| 298 |
+
"learning_rate": 7.42304939155333e-05,
|
| 299 |
+
"loss": 1.0348,
|
| 300 |
+
"step": 460
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.94,
|
| 304 |
+
"learning_rate": 7.351467430207589e-05,
|
| 305 |
+
"loss": 0.9361,
|
| 306 |
+
"step": 470
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.96,
|
| 310 |
+
"learning_rate": 7.279885468861847e-05,
|
| 311 |
+
"loss": 0.8108,
|
| 312 |
+
"step": 480
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.98,
|
| 316 |
+
"learning_rate": 7.208303507516106e-05,
|
| 317 |
+
"loss": 0.677,
|
| 318 |
+
"step": 490
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.0,
|
| 322 |
+
"learning_rate": 7.136721546170365e-05,
|
| 323 |
+
"loss": 0.6402,
|
| 324 |
+
"step": 500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.02,
|
| 328 |
+
"learning_rate": 7.065139584824624e-05,
|
| 329 |
+
"loss": 1.0431,
|
| 330 |
+
"step": 510
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.04,
|
| 334 |
+
"learning_rate": 6.993557623478884e-05,
|
| 335 |
+
"loss": 0.9406,
|
| 336 |
+
"step": 520
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.06,
|
| 340 |
+
"learning_rate": 6.921975662133143e-05,
|
| 341 |
+
"loss": 0.7971,
|
| 342 |
+
"step": 530
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.08,
|
| 346 |
+
"learning_rate": 6.850393700787402e-05,
|
| 347 |
+
"loss": 0.6652,
|
| 348 |
+
"step": 540
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.1,
|
| 352 |
+
"learning_rate": 6.778811739441661e-05,
|
| 353 |
+
"loss": 0.651,
|
| 354 |
+
"step": 550
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.12,
|
| 358 |
+
"learning_rate": 6.70722977809592e-05,
|
| 359 |
+
"loss": 1.0322,
|
| 360 |
+
"step": 560
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.14,
|
| 364 |
+
"learning_rate": 6.63564781675018e-05,
|
| 365 |
+
"loss": 0.9425,
|
| 366 |
+
"step": 570
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.16,
|
| 370 |
+
"learning_rate": 6.564065855404439e-05,
|
| 371 |
+
"loss": 0.8085,
|
| 372 |
+
"step": 580
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 1.18,
|
| 376 |
+
"learning_rate": 6.492483894058698e-05,
|
| 377 |
+
"loss": 0.6654,
|
| 378 |
+
"step": 590
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.2,
|
| 382 |
+
"learning_rate": 6.420901932712957e-05,
|
| 383 |
+
"loss": 0.6493,
|
| 384 |
+
"step": 600
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.2,
|
| 388 |
+
"eval_loss": 0.9059903025627136,
|
| 389 |
+
"eval_runtime": 133.8487,
|
| 390 |
+
"eval_samples_per_second": 14.942,
|
| 391 |
+
"eval_steps_per_second": 1.868,
|
| 392 |
+
"step": 600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.22,
|
| 396 |
+
"learning_rate": 6.349319971367216e-05,
|
| 397 |
+
"loss": 1.0063,
|
| 398 |
+
"step": 610
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.24,
|
| 402 |
+
"learning_rate": 6.277738010021474e-05,
|
| 403 |
+
"loss": 0.9179,
|
| 404 |
+
"step": 620
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.26,
|
| 408 |
+
"learning_rate": 6.206156048675734e-05,
|
| 409 |
+
"loss": 0.797,
|
| 410 |
+
"step": 630
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.28,
|
| 414 |
+
"learning_rate": 6.134574087329993e-05,
|
| 415 |
+
"loss": 0.6743,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.3,
|
| 420 |
+
"learning_rate": 6.0629921259842526e-05,
|
| 421 |
+
"loss": 0.6468,
|
| 422 |
+
"step": 650
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.32,
|
| 426 |
+
"learning_rate": 5.991410164638511e-05,
|
| 427 |
+
"loss": 1.0237,
|
| 428 |
+
"step": 660
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.34,
|
| 432 |
+
"learning_rate": 5.919828203292771e-05,
|
| 433 |
+
"loss": 0.9141,
|
| 434 |
+
"step": 670
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.36,
|
| 438 |
+
"learning_rate": 5.8482462419470296e-05,
|
| 439 |
+
"loss": 0.7917,
|
| 440 |
+
"step": 680
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.38,
|
| 444 |
+
"learning_rate": 5.776664280601288e-05,
|
| 445 |
+
"loss": 0.6499,
|
| 446 |
+
"step": 690
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.4,
|
| 450 |
+
"learning_rate": 5.705082319255548e-05,
|
| 451 |
+
"loss": 0.6481,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.42,
|
| 456 |
+
"learning_rate": 5.6335003579098066e-05,
|
| 457 |
+
"loss": 1.0164,
|
| 458 |
+
"step": 710
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.44,
|
| 462 |
+
"learning_rate": 5.5619183965640665e-05,
|
| 463 |
+
"loss": 0.9252,
|
| 464 |
+
"step": 720
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.46,
|
| 468 |
+
"learning_rate": 5.490336435218325e-05,
|
| 469 |
+
"loss": 0.7942,
|
| 470 |
+
"step": 730
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.48,
|
| 474 |
+
"learning_rate": 5.418754473872585e-05,
|
| 475 |
+
"loss": 0.6517,
|
| 476 |
+
"step": 740
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.5,
|
| 480 |
+
"learning_rate": 5.3471725125268434e-05,
|
| 481 |
+
"loss": 0.642,
|
| 482 |
+
"step": 750
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.52,
|
| 486 |
+
"learning_rate": 5.275590551181102e-05,
|
| 487 |
+
"loss": 1.0249,
|
| 488 |
+
"step": 760
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.54,
|
| 492 |
+
"learning_rate": 5.204008589835362e-05,
|
| 493 |
+
"loss": 0.9345,
|
| 494 |
+
"step": 770
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.56,
|
| 498 |
+
"learning_rate": 5.1324266284896204e-05,
|
| 499 |
+
"loss": 0.7841,
|
| 500 |
+
"step": 780
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.58,
|
| 504 |
+
"learning_rate": 5.06084466714388e-05,
|
| 505 |
+
"loss": 0.6512,
|
| 506 |
+
"step": 790
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6,
|
| 510 |
+
"learning_rate": 4.989262705798139e-05,
|
| 511 |
+
"loss": 0.6472,
|
| 512 |
+
"step": 800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.6,
|
| 516 |
+
"eval_loss": 0.8923133611679077,
|
| 517 |
+
"eval_runtime": 133.1272,
|
| 518 |
+
"eval_samples_per_second": 15.023,
|
| 519 |
+
"eval_steps_per_second": 1.878,
|
| 520 |
+
"step": 800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.62,
|
| 524 |
+
"learning_rate": 4.917680744452398e-05,
|
| 525 |
+
"loss": 1.0134,
|
| 526 |
+
"step": 810
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.64,
|
| 530 |
+
"learning_rate": 4.846098783106657e-05,
|
| 531 |
+
"loss": 0.9177,
|
| 532 |
+
"step": 820
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.66,
|
| 536 |
+
"learning_rate": 4.7745168217609165e-05,
|
| 537 |
+
"loss": 0.786,
|
| 538 |
+
"step": 830
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.68,
|
| 542 |
+
"learning_rate": 4.702934860415176e-05,
|
| 543 |
+
"loss": 0.6569,
|
| 544 |
+
"step": 840
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.7,
|
| 548 |
+
"learning_rate": 4.631352899069435e-05,
|
| 549 |
+
"loss": 0.6258,
|
| 550 |
+
"step": 850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.72,
|
| 554 |
+
"learning_rate": 4.559770937723694e-05,
|
| 555 |
+
"loss": 1.0336,
|
| 556 |
+
"step": 860
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.74,
|
| 560 |
+
"learning_rate": 4.488188976377953e-05,
|
| 561 |
+
"loss": 0.9346,
|
| 562 |
+
"step": 870
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.76,
|
| 566 |
+
"learning_rate": 4.416607015032212e-05,
|
| 567 |
+
"loss": 0.7893,
|
| 568 |
+
"step": 880
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.78,
|
| 572 |
+
"learning_rate": 4.345025053686471e-05,
|
| 573 |
+
"loss": 0.6585,
|
| 574 |
+
"step": 890
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.8,
|
| 578 |
+
"learning_rate": 4.27344309234073e-05,
|
| 579 |
+
"loss": 0.6222,
|
| 580 |
+
"step": 900
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.82,
|
| 584 |
+
"learning_rate": 4.2018611309949896e-05,
|
| 585 |
+
"loss": 1.0196,
|
| 586 |
+
"step": 910
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.84,
|
| 590 |
+
"learning_rate": 4.130279169649249e-05,
|
| 591 |
+
"loss": 0.9398,
|
| 592 |
+
"step": 920
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.86,
|
| 596 |
+
"learning_rate": 4.058697208303508e-05,
|
| 597 |
+
"loss": 0.8012,
|
| 598 |
+
"step": 930
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.88,
|
| 602 |
+
"learning_rate": 3.9871152469577665e-05,
|
| 603 |
+
"loss": 0.6599,
|
| 604 |
+
"step": 940
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.9,
|
| 608 |
+
"learning_rate": 3.915533285612026e-05,
|
| 609 |
+
"loss": 0.6337,
|
| 610 |
+
"step": 950
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.92,
|
| 614 |
+
"learning_rate": 3.843951324266285e-05,
|
| 615 |
+
"loss": 1.0181,
|
| 616 |
+
"step": 960
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.94,
|
| 620 |
+
"learning_rate": 3.772369362920544e-05,
|
| 621 |
+
"loss": 0.9119,
|
| 622 |
+
"step": 970
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.96,
|
| 626 |
+
"learning_rate": 3.7007874015748034e-05,
|
| 627 |
+
"loss": 0.7974,
|
| 628 |
+
"step": 980
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.98,
|
| 632 |
+
"learning_rate": 3.6292054402290626e-05,
|
| 633 |
+
"loss": 0.6506,
|
| 634 |
+
"step": 990
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.0,
|
| 638 |
+
"learning_rate": 3.557623478883322e-05,
|
| 639 |
+
"loss": 0.6552,
|
| 640 |
+
"step": 1000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.0,
|
| 644 |
+
"eval_loss": 0.883411169052124,
|
| 645 |
+
"eval_runtime": 133.1959,
|
| 646 |
+
"eval_samples_per_second": 15.015,
|
| 647 |
+
"eval_steps_per_second": 1.877,
|
| 648 |
+
"step": 1000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.02,
|
| 652 |
+
"learning_rate": 3.4860415175375804e-05,
|
| 653 |
+
"loss": 1.0039,
|
| 654 |
+
"step": 1010
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 2.04,
|
| 658 |
+
"learning_rate": 3.4144595561918396e-05,
|
| 659 |
+
"loss": 0.9259,
|
| 660 |
+
"step": 1020
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.06,
|
| 664 |
+
"learning_rate": 3.342877594846099e-05,
|
| 665 |
+
"loss": 0.7858,
|
| 666 |
+
"step": 1030
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 2.08,
|
| 670 |
+
"learning_rate": 3.271295633500358e-05,
|
| 671 |
+
"loss": 0.649,
|
| 672 |
+
"step": 1040
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.1,
|
| 676 |
+
"learning_rate": 3.199713672154617e-05,
|
| 677 |
+
"loss": 0.6557,
|
| 678 |
+
"step": 1050
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.12,
|
| 682 |
+
"learning_rate": 3.1281317108088764e-05,
|
| 683 |
+
"loss": 1.0129,
|
| 684 |
+
"step": 1060
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.14,
|
| 688 |
+
"learning_rate": 3.0565497494631357e-05,
|
| 689 |
+
"loss": 0.901,
|
| 690 |
+
"step": 1070
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.16,
|
| 694 |
+
"learning_rate": 2.9849677881173942e-05,
|
| 695 |
+
"loss": 0.7697,
|
| 696 |
+
"step": 1080
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 2.18,
|
| 700 |
+
"learning_rate": 2.9133858267716534e-05,
|
| 701 |
+
"loss": 0.6419,
|
| 702 |
+
"step": 1090
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.2,
|
| 706 |
+
"learning_rate": 2.8418038654259126e-05,
|
| 707 |
+
"loss": 0.6546,
|
| 708 |
+
"step": 1100
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 2.22,
|
| 712 |
+
"learning_rate": 2.770221904080172e-05,
|
| 713 |
+
"loss": 1.009,
|
| 714 |
+
"step": 1110
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.24,
|
| 718 |
+
"learning_rate": 2.698639942734431e-05,
|
| 719 |
+
"loss": 0.9193,
|
| 720 |
+
"step": 1120
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.26,
|
| 724 |
+
"learning_rate": 2.6270579813886903e-05,
|
| 725 |
+
"loss": 0.7808,
|
| 726 |
+
"step": 1130
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.28,
|
| 730 |
+
"learning_rate": 2.5554760200429495e-05,
|
| 731 |
+
"loss": 0.6449,
|
| 732 |
+
"step": 1140
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.3,
|
| 736 |
+
"learning_rate": 2.4838940586972084e-05,
|
| 737 |
+
"loss": 0.6551,
|
| 738 |
+
"step": 1150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.32,
|
| 742 |
+
"learning_rate": 2.4123120973514676e-05,
|
| 743 |
+
"loss": 1.0013,
|
| 744 |
+
"step": 1160
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.34,
|
| 748 |
+
"learning_rate": 2.3407301360057265e-05,
|
| 749 |
+
"loss": 0.9067,
|
| 750 |
+
"step": 1170
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.36,
|
| 754 |
+
"learning_rate": 2.2691481746599857e-05,
|
| 755 |
+
"loss": 0.7739,
|
| 756 |
+
"step": 1180
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 2.38,
|
| 760 |
+
"learning_rate": 2.197566213314245e-05,
|
| 761 |
+
"loss": 0.6502,
|
| 762 |
+
"step": 1190
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.4,
|
| 766 |
+
"learning_rate": 2.125984251968504e-05,
|
| 767 |
+
"loss": 0.6557,
|
| 768 |
+
"step": 1200
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.4,
|
| 772 |
+
"eval_loss": 0.8776447176933289,
|
| 773 |
+
"eval_runtime": 133.098,
|
| 774 |
+
"eval_samples_per_second": 15.027,
|
| 775 |
+
"eval_steps_per_second": 1.878,
|
| 776 |
+
"step": 1200
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.42,
|
| 780 |
+
"learning_rate": 2.054402290622763e-05,
|
| 781 |
+
"loss": 0.9857,
|
| 782 |
+
"step": 1210
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.44,
|
| 786 |
+
"learning_rate": 1.9828203292770222e-05,
|
| 787 |
+
"loss": 0.8983,
|
| 788 |
+
"step": 1220
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.46,
|
| 792 |
+
"learning_rate": 1.9112383679312814e-05,
|
| 793 |
+
"loss": 0.7759,
|
| 794 |
+
"step": 1230
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.48,
|
| 798 |
+
"learning_rate": 1.8396564065855406e-05,
|
| 799 |
+
"loss": 0.65,
|
| 800 |
+
"step": 1240
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.5,
|
| 804 |
+
"learning_rate": 1.7680744452397995e-05,
|
| 805 |
+
"loss": 0.6538,
|
| 806 |
+
"step": 1250
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.52,
|
| 810 |
+
"learning_rate": 1.6964924838940587e-05,
|
| 811 |
+
"loss": 0.991,
|
| 812 |
+
"step": 1260
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.54,
|
| 816 |
+
"learning_rate": 1.624910522548318e-05,
|
| 817 |
+
"loss": 0.9044,
|
| 818 |
+
"step": 1270
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.56,
|
| 822 |
+
"learning_rate": 1.553328561202577e-05,
|
| 823 |
+
"loss": 0.7839,
|
| 824 |
+
"step": 1280
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.58,
|
| 828 |
+
"learning_rate": 1.481746599856836e-05,
|
| 829 |
+
"loss": 0.6432,
|
| 830 |
+
"step": 1290
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.6,
|
| 834 |
+
"learning_rate": 1.4101646385110953e-05,
|
| 835 |
+
"loss": 0.656,
|
| 836 |
+
"step": 1300
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.62,
|
| 840 |
+
"learning_rate": 1.3385826771653545e-05,
|
| 841 |
+
"loss": 0.9855,
|
| 842 |
+
"step": 1310
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.64,
|
| 846 |
+
"learning_rate": 1.2670007158196134e-05,
|
| 847 |
+
"loss": 0.9121,
|
| 848 |
+
"step": 1320
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.66,
|
| 852 |
+
"learning_rate": 1.1954187544738726e-05,
|
| 853 |
+
"loss": 0.7801,
|
| 854 |
+
"step": 1330
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.68,
|
| 858 |
+
"learning_rate": 1.1238367931281316e-05,
|
| 859 |
+
"loss": 0.6426,
|
| 860 |
+
"step": 1340
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.7,
|
| 864 |
+
"learning_rate": 1.0522548317823908e-05,
|
| 865 |
+
"loss": 0.6474,
|
| 866 |
+
"step": 1350
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.72,
|
| 870 |
+
"learning_rate": 9.806728704366499e-06,
|
| 871 |
+
"loss": 1.009,
|
| 872 |
+
"step": 1360
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.74,
|
| 876 |
+
"learning_rate": 9.090909090909091e-06,
|
| 877 |
+
"loss": 0.9203,
|
| 878 |
+
"step": 1370
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.76,
|
| 882 |
+
"learning_rate": 8.375089477451682e-06,
|
| 883 |
+
"loss": 0.7858,
|
| 884 |
+
"step": 1380
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.78,
|
| 888 |
+
"learning_rate": 7.659269863994274e-06,
|
| 889 |
+
"loss": 0.644,
|
| 890 |
+
"step": 1390
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.8,
|
| 894 |
+
"learning_rate": 6.943450250536865e-06,
|
| 895 |
+
"loss": 0.6542,
|
| 896 |
+
"step": 1400
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.8,
|
| 900 |
+
"eval_loss": 0.8746508359909058,
|
| 901 |
+
"eval_runtime": 133.0912,
|
| 902 |
+
"eval_samples_per_second": 15.027,
|
| 903 |
+
"eval_steps_per_second": 1.878,
|
| 904 |
+
"step": 1400
|
| 905 |
+
}
|
| 906 |
+
],
|
| 907 |
+
"max_steps": 1497,
|
| 908 |
+
"num_train_epochs": 3,
|
| 909 |
+
"total_flos": 7.421122325073101e+17,
|
| 910 |
+
"trial_name": null,
|
| 911 |
+
"trial_params": null
|
| 912 |
+
}
|
lora-alpaca/checkpoint-1400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3365a9fddc28ac19526733cfa419ff1af8b6a9569c78e529ea4a0e67de3e3a30
|
| 3 |
+
size 3515
|