Upload 8 files
Browse files- adapter_config.json +37 -0
- adapter_model.safetensors +3 -0
- additional_config.json +1 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +577 -0
- training_args.bin +3 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "Qwen2.5-14B-Instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"eva_config": null,
|
| 7 |
+
"exclude_modules": null,
|
| 8 |
+
"fan_in_fan_out": false,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": true,
|
| 11 |
+
"layer_replication": null,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 32,
|
| 16 |
+
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.0,
|
| 18 |
+
"megatron_config": null,
|
| 19 |
+
"megatron_core": "megatron.core",
|
| 20 |
+
"modules_to_save": null,
|
| 21 |
+
"peft_type": "LORA",
|
| 22 |
+
"r": 16,
|
| 23 |
+
"rank_pattern": {},
|
| 24 |
+
"revision": null,
|
| 25 |
+
"target_modules": [
|
| 26 |
+
"up_proj",
|
| 27 |
+
"gate_proj",
|
| 28 |
+
"o_proj",
|
| 29 |
+
"k_proj",
|
| 30 |
+
"v_proj",
|
| 31 |
+
"q_proj",
|
| 32 |
+
"down_proj"
|
| 33 |
+
],
|
| 34 |
+
"task_type": "CAUSAL_LM",
|
| 35 |
+
"use_dora": false,
|
| 36 |
+
"use_rslora": false
|
| 37 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3edbcc01d656f24ff31bf7c8e46431c79642b4f4665ca85ef0164a2b426a0016
|
| 3 |
+
size 275341720
|
additional_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bd9af23ddd12b10d8c87087f93cabaca9fc51a3c2566a3c823fd1b1c7cb472b
|
| 3 |
+
size 551070514
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d138cfe3a4adf21f048848ee35837c9a757a0a3616ff7adbb45b69aac247435
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5bfdcce041c9ea47daf7124b9987c48c3cd52a01a7502e37b705c7697288a10
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,577 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.57263505,
|
| 3 |
+
"best_model_checkpoint": "/export/home2/zli/kc/mm_rag/Qwen2.5-14B-Instruct_lora/checkpoint-256",
|
| 4 |
+
"epoch": 0.9978075517661389,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 256,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00389768574908648,
|
| 13 |
+
"grad_norm": 0.3756321966648102,
|
| 14 |
+
"learning_rate": 7.692307692307694e-06,
|
| 15 |
+
"loss": 0.760678231716156,
|
| 16 |
+
"memory(GiB)": 64.79,
|
| 17 |
+
"step": 1,
|
| 18 |
+
"token_acc": 0.794377180381695,
|
| 19 |
+
"train_speed(iter/s)": 0.008942
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.0194884287454324,
|
| 23 |
+
"grad_norm": 0.4006075859069824,
|
| 24 |
+
"learning_rate": 3.846153846153846e-05,
|
| 25 |
+
"loss": 0.8511521816253662,
|
| 26 |
+
"memory(GiB)": 69.21,
|
| 27 |
+
"step": 5,
|
| 28 |
+
"token_acc": 0.7628462851792697,
|
| 29 |
+
"train_speed(iter/s)": 0.009262
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.0389768574908648,
|
| 33 |
+
"grad_norm": 0.40541699528694153,
|
| 34 |
+
"learning_rate": 7.692307692307693e-05,
|
| 35 |
+
"loss": 0.8799286842346191,
|
| 36 |
+
"memory(GiB)": 69.21,
|
| 37 |
+
"step": 10,
|
| 38 |
+
"token_acc": 0.755640420079393,
|
| 39 |
+
"train_speed(iter/s)": 0.009459
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.058465286236297195,
|
| 43 |
+
"grad_norm": 0.28694042563438416,
|
| 44 |
+
"learning_rate": 9.998328666948438e-05,
|
| 45 |
+
"loss": 0.7268682479858398,
|
| 46 |
+
"memory(GiB)": 69.21,
|
| 47 |
+
"step": 15,
|
| 48 |
+
"token_acc": 0.7857048253940223,
|
| 49 |
+
"train_speed(iter/s)": 0.009584
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.0779537149817296,
|
| 53 |
+
"grad_norm": 0.29239076375961304,
|
| 54 |
+
"learning_rate": 9.979538999730047e-05,
|
| 55 |
+
"loss": 0.6990458011627197,
|
| 56 |
+
"memory(GiB)": 69.21,
|
| 57 |
+
"step": 20,
|
| 58 |
+
"token_acc": 0.7923689929968607,
|
| 59 |
+
"train_speed(iter/s)": 0.009633
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09744214372716199,
|
| 63 |
+
"grad_norm": 0.25992509722709656,
|
| 64 |
+
"learning_rate": 9.939949247384046e-05,
|
| 65 |
+
"loss": 0.7058870792388916,
|
| 66 |
+
"memory(GiB)": 71.47,
|
| 67 |
+
"step": 25,
|
| 68 |
+
"token_acc": 0.7879081525173794,
|
| 69 |
+
"train_speed(iter/s)": 0.009549
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.11693057247259439,
|
| 73 |
+
"grad_norm": 0.1924816370010376,
|
| 74 |
+
"learning_rate": 9.879724780684519e-05,
|
| 75 |
+
"loss": 0.7370085716247559,
|
| 76 |
+
"memory(GiB)": 71.47,
|
| 77 |
+
"step": 30,
|
| 78 |
+
"token_acc": 0.7818104986383698,
|
| 79 |
+
"train_speed(iter/s)": 0.009632
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1364190012180268,
|
| 83 |
+
"grad_norm": 0.2375946342945099,
|
| 84 |
+
"learning_rate": 9.799117163889559e-05,
|
| 85 |
+
"loss": 0.6951110363006592,
|
| 86 |
+
"memory(GiB)": 71.47,
|
| 87 |
+
"step": 35,
|
| 88 |
+
"token_acc": 0.7915360501567398,
|
| 89 |
+
"train_speed(iter/s)": 0.009685
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.1559074299634592,
|
| 93 |
+
"grad_norm": 0.19938509166240692,
|
| 94 |
+
"learning_rate": 9.698463103929542e-05,
|
| 95 |
+
"loss": 0.634189510345459,
|
| 96 |
+
"memory(GiB)": 71.47,
|
| 97 |
+
"step": 40,
|
| 98 |
+
"token_acc": 0.8057028291378926,
|
| 99 |
+
"train_speed(iter/s)": 0.009727
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.1753958587088916,
|
| 103 |
+
"grad_norm": 0.2671877145767212,
|
| 104 |
+
"learning_rate": 9.57818304394503e-05,
|
| 105 |
+
"loss": 0.6699491500854492,
|
| 106 |
+
"memory(GiB)": 71.47,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"token_acc": 0.8010264675941635,
|
| 109 |
+
"train_speed(iter/s)": 0.009776
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.19488428745432398,
|
| 113 |
+
"grad_norm": 0.3463566303253174,
|
| 114 |
+
"learning_rate": 9.438779407049281e-05,
|
| 115 |
+
"loss": 0.6760294437408447,
|
| 116 |
+
"memory(GiB)": 71.47,
|
| 117 |
+
"step": 50,
|
| 118 |
+
"token_acc": 0.793778801843318,
|
| 119 |
+
"train_speed(iter/s)": 0.009808
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.2143727161997564,
|
| 123 |
+
"grad_norm": 0.22926518321037292,
|
| 124 |
+
"learning_rate": 9.280834497651334e-05,
|
| 125 |
+
"loss": 0.6952155113220215,
|
| 126 |
+
"memory(GiB)": 71.47,
|
| 127 |
+
"step": 55,
|
| 128 |
+
"token_acc": 0.7856186209763081,
|
| 129 |
+
"train_speed(iter/s)": 0.009819
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.23386114494518878,
|
| 133 |
+
"grad_norm": 0.26200923323631287,
|
| 134 |
+
"learning_rate": 9.105008069106093e-05,
|
| 135 |
+
"loss": 0.6922572135925293,
|
| 136 |
+
"memory(GiB)": 71.47,
|
| 137 |
+
"step": 60,
|
| 138 |
+
"token_acc": 0.784200469035004,
|
| 139 |
+
"train_speed(iter/s)": 0.009839
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.25334957369062117,
|
| 143 |
+
"grad_norm": 0.2608520984649658,
|
| 144 |
+
"learning_rate": 8.912034567851599e-05,
|
| 145 |
+
"loss": 0.6162422180175782,
|
| 146 |
+
"memory(GiB)": 71.47,
|
| 147 |
+
"step": 65,
|
| 148 |
+
"token_acc": 0.8122029543994862,
|
| 149 |
+
"train_speed(iter/s)": 0.009864
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2728380024360536,
|
| 153 |
+
"grad_norm": 0.23958414793014526,
|
| 154 |
+
"learning_rate": 8.702720065545024e-05,
|
| 155 |
+
"loss": 0.6201987266540527,
|
| 156 |
+
"memory(GiB)": 71.47,
|
| 157 |
+
"step": 70,
|
| 158 |
+
"token_acc": 0.807689317368994,
|
| 159 |
+
"train_speed(iter/s)": 0.00982
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.292326431181486,
|
| 163 |
+
"grad_norm": 0.24707700312137604,
|
| 164 |
+
"learning_rate": 8.47793889201221e-05,
|
| 165 |
+
"loss": 0.6751769542694092,
|
| 166 |
+
"memory(GiB)": 71.47,
|
| 167 |
+
"step": 75,
|
| 168 |
+
"token_acc": 0.7933965994464215,
|
| 169 |
+
"train_speed(iter/s)": 0.009843
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.3118148599269184,
|
| 173 |
+
"grad_norm": 0.280536413192749,
|
| 174 |
+
"learning_rate": 8.238629983075294e-05,
|
| 175 |
+
"loss": 0.6537846088409424,
|
| 176 |
+
"memory(GiB)": 71.47,
|
| 177 |
+
"step": 80,
|
| 178 |
+
"token_acc": 0.800564772326156,
|
| 179 |
+
"train_speed(iter/s)": 0.00983
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.3313032886723508,
|
| 183 |
+
"grad_norm": 0.29046493768692017,
|
| 184 |
+
"learning_rate": 7.985792958513931e-05,
|
| 185 |
+
"loss": 0.6765205383300781,
|
| 186 |
+
"memory(GiB)": 71.47,
|
| 187 |
+
"step": 85,
|
| 188 |
+
"token_acc": 0.7913814331445457,
|
| 189 |
+
"train_speed(iter/s)": 0.009828
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.3507917174177832,
|
| 193 |
+
"grad_norm": 0.2879982888698578,
|
| 194 |
+
"learning_rate": 7.720483946542914e-05,
|
| 195 |
+
"loss": 0.6516193866729736,
|
| 196 |
+
"memory(GiB)": 71.47,
|
| 197 |
+
"step": 90,
|
| 198 |
+
"token_acc": 0.7983723394009439,
|
| 199 |
+
"train_speed(iter/s)": 0.009817
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3702801461632156,
|
| 203 |
+
"grad_norm": 0.28098541498184204,
|
| 204 |
+
"learning_rate": 7.443811172247821e-05,
|
| 205 |
+
"loss": 0.6455563545227051,
|
| 206 |
+
"memory(GiB)": 71.47,
|
| 207 |
+
"step": 95,
|
| 208 |
+
"token_acc": 0.8004983012457532,
|
| 209 |
+
"train_speed(iter/s)": 0.009833
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.38976857490864797,
|
| 213 |
+
"grad_norm": 0.2407151460647583,
|
| 214 |
+
"learning_rate": 7.156930328406268e-05,
|
| 215 |
+
"loss": 0.6356846332550049,
|
| 216 |
+
"memory(GiB)": 71.47,
|
| 217 |
+
"step": 100,
|
| 218 |
+
"token_acc": 0.8042028604018602,
|
| 219 |
+
"train_speed(iter/s)": 0.009804
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.38976857490864797,
|
| 223 |
+
"eval_loss": 0.598721981048584,
|
| 224 |
+
"eval_runtime": 15.1098,
|
| 225 |
+
"eval_samples_per_second": 0.265,
|
| 226 |
+
"eval_steps_per_second": 0.132,
|
| 227 |
+
"step": 100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4092570036540804,
|
| 231 |
+
"grad_norm": 0.2836802899837494,
|
| 232 |
+
"learning_rate": 6.861039748031351e-05,
|
| 233 |
+
"loss": 0.6598537921905517,
|
| 234 |
+
"memory(GiB)": 77.63,
|
| 235 |
+
"step": 105,
|
| 236 |
+
"token_acc": 0.7957189390414147,
|
| 237 |
+
"train_speed(iter/s)": 0.009789
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.4287454323995128,
|
| 241 |
+
"grad_norm": 0.26613113284111023,
|
| 242 |
+
"learning_rate": 6.557375398802123e-05,
|
| 243 |
+
"loss": 0.6290350437164307,
|
| 244 |
+
"memory(GiB)": 77.63,
|
| 245 |
+
"step": 110,
|
| 246 |
+
"token_acc": 0.8080845540515483,
|
| 247 |
+
"train_speed(iter/s)": 0.009743
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.4482338611449452,
|
| 251 |
+
"grad_norm": 0.3952692747116089,
|
| 252 |
+
"learning_rate": 6.247205720289907e-05,
|
| 253 |
+
"loss": 0.6128222942352295,
|
| 254 |
+
"memory(GiB)": 77.63,
|
| 255 |
+
"step": 115,
|
| 256 |
+
"token_acc": 0.8121368342872122,
|
| 257 |
+
"train_speed(iter/s)": 0.009719
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.46772228989037756,
|
| 261 |
+
"grad_norm": 0.29315847158432007,
|
| 262 |
+
"learning_rate": 5.9318263255459116e-05,
|
| 263 |
+
"loss": 0.6240291595458984,
|
| 264 |
+
"memory(GiB)": 77.63,
|
| 265 |
+
"step": 120,
|
| 266 |
+
"token_acc": 0.8027493895269965,
|
| 267 |
+
"train_speed(iter/s)": 0.009694
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.48721071863581,
|
| 271 |
+
"grad_norm": 0.2671997547149658,
|
| 272 |
+
"learning_rate": 5.6125545891822274e-05,
|
| 273 |
+
"loss": 0.6649426460266114,
|
| 274 |
+
"memory(GiB)": 77.63,
|
| 275 |
+
"step": 125,
|
| 276 |
+
"token_acc": 0.7922591927838373,
|
| 277 |
+
"train_speed(iter/s)": 0.009676
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.5066991473812423,
|
| 281 |
+
"grad_norm": 0.26974523067474365,
|
| 282 |
+
"learning_rate": 5.290724144552379e-05,
|
| 283 |
+
"loss": 0.6172929763793945,
|
| 284 |
+
"memory(GiB)": 77.63,
|
| 285 |
+
"step": 130,
|
| 286 |
+
"token_acc": 0.8085726351351351,
|
| 287 |
+
"train_speed(iter/s)": 0.009667
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.5261875761266748,
|
| 291 |
+
"grad_norm": 0.2798590660095215,
|
| 292 |
+
"learning_rate": 4.967679313017303e-05,
|
| 293 |
+
"loss": 0.6145929336547852,
|
| 294 |
+
"memory(GiB)": 77.63,
|
| 295 |
+
"step": 135,
|
| 296 |
+
"token_acc": 0.8118081180811808,
|
| 297 |
+
"train_speed(iter/s)": 0.009683
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5456760048721072,
|
| 301 |
+
"grad_norm": 0.27218812704086304,
|
| 302 |
+
"learning_rate": 4.6447694885663514e-05,
|
| 303 |
+
"loss": 0.618541955947876,
|
| 304 |
+
"memory(GiB)": 77.63,
|
| 305 |
+
"step": 140,
|
| 306 |
+
"token_acc": 0.8050674649659422,
|
| 307 |
+
"train_speed(iter/s)": 0.009706
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.5651644336175395,
|
| 311 |
+
"grad_norm": 0.29498329758644104,
|
| 312 |
+
"learning_rate": 4.323343501249346e-05,
|
| 313 |
+
"loss": 0.6341890335083008,
|
| 314 |
+
"memory(GiB)": 77.63,
|
| 315 |
+
"step": 145,
|
| 316 |
+
"token_acc": 0.801550316061459,
|
| 317 |
+
"train_speed(iter/s)": 0.009735
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.584652862362972,
|
| 321 |
+
"grad_norm": 0.2799496352672577,
|
| 322 |
+
"learning_rate": 4.004743982964298e-05,
|
| 323 |
+
"loss": 0.6000133037567139,
|
| 324 |
+
"memory(GiB)": 77.63,
|
| 325 |
+
"step": 150,
|
| 326 |
+
"token_acc": 0.8125188083057479,
|
| 327 |
+
"train_speed(iter/s)": 0.009742
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.6041412911084044,
|
| 331 |
+
"grad_norm": 0.28971585631370544,
|
| 332 |
+
"learning_rate": 3.6903017591354706e-05,
|
| 333 |
+
"loss": 0.6476753711700439,
|
| 334 |
+
"memory(GiB)": 77.63,
|
| 335 |
+
"step": 155,
|
| 336 |
+
"token_acc": 0.8055159053833605,
|
| 337 |
+
"train_speed(iter/s)": 0.009734
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.6236297198538368,
|
| 341 |
+
"grad_norm": 0.2888241708278656,
|
| 342 |
+
"learning_rate": 3.381330289708396e-05,
|
| 343 |
+
"loss": 0.6604040622711181,
|
| 344 |
+
"memory(GiB)": 77.63,
|
| 345 |
+
"step": 160,
|
| 346 |
+
"token_acc": 0.7972270363951474,
|
| 347 |
+
"train_speed(iter/s)": 0.009743
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.6431181485992692,
|
| 351 |
+
"grad_norm": 0.31141144037246704,
|
| 352 |
+
"learning_rate": 3.079120182682412e-05,
|
| 353 |
+
"loss": 0.5943418025970459,
|
| 354 |
+
"memory(GiB)": 77.63,
|
| 355 |
+
"step": 165,
|
| 356 |
+
"token_acc": 0.8125655328607977,
|
| 357 |
+
"train_speed(iter/s)": 0.009733
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.6626065773447016,
|
| 361 |
+
"grad_norm": 0.28593260049819946,
|
| 362 |
+
"learning_rate": 2.7849338030983257e-05,
|
| 363 |
+
"loss": 0.6077968120574951,
|
| 364 |
+
"memory(GiB)": 77.63,
|
| 365 |
+
"step": 170,
|
| 366 |
+
"token_acc": 0.811592125382263,
|
| 367 |
+
"train_speed(iter/s)": 0.009731
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.682095006090134,
|
| 371 |
+
"grad_norm": 0.3223704993724823,
|
| 372 |
+
"learning_rate": 2.500000000000001e-05,
|
| 373 |
+
"loss": 0.6201191425323487,
|
| 374 |
+
"memory(GiB)": 77.63,
|
| 375 |
+
"step": 175,
|
| 376 |
+
"token_acc": 0.806087783666259,
|
| 377 |
+
"train_speed(iter/s)": 0.009728
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.7015834348355664,
|
| 381 |
+
"grad_norm": 0.3051086962223053,
|
| 382 |
+
"learning_rate": 2.225508973396016e-05,
|
| 383 |
+
"loss": 0.6003672122955322,
|
| 384 |
+
"memory(GiB)": 77.63,
|
| 385 |
+
"step": 180,
|
| 386 |
+
"token_acc": 0.8096822043421585,
|
| 387 |
+
"train_speed(iter/s)": 0.009738
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.7210718635809987,
|
| 391 |
+
"grad_norm": 0.33248814940452576,
|
| 392 |
+
"learning_rate": 1.9626073026625818e-05,
|
| 393 |
+
"loss": 0.6649184226989746,
|
| 394 |
+
"memory(GiB)": 77.63,
|
| 395 |
+
"step": 185,
|
| 396 |
+
"token_acc": 0.7927942418271744,
|
| 397 |
+
"train_speed(iter/s)": 0.009754
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.7405602923264312,
|
| 401 |
+
"grad_norm": 0.32766395807266235,
|
| 402 |
+
"learning_rate": 1.7123931571546827e-05,
|
| 403 |
+
"loss": 0.620220422744751,
|
| 404 |
+
"memory(GiB)": 77.63,
|
| 405 |
+
"step": 190,
|
| 406 |
+
"token_acc": 0.8047315233083407,
|
| 407 |
+
"train_speed(iter/s)": 0.009761
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.7600487210718636,
|
| 411 |
+
"grad_norm": 0.3414992392063141,
|
| 412 |
+
"learning_rate": 1.4759117090312197e-05,
|
| 413 |
+
"loss": 0.622031831741333,
|
| 414 |
+
"memory(GiB)": 77.63,
|
| 415 |
+
"step": 195,
|
| 416 |
+
"token_acc": 0.807596541264883,
|
| 417 |
+
"train_speed(iter/s)": 0.009765
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.7795371498172959,
|
| 421 |
+
"grad_norm": 0.32977092266082764,
|
| 422 |
+
"learning_rate": 1.25415076745532e-05,
|
| 423 |
+
"loss": 0.6251283168792725,
|
| 424 |
+
"memory(GiB)": 77.63,
|
| 425 |
+
"step": 200,
|
| 426 |
+
"token_acc": 0.8072234762979684,
|
| 427 |
+
"train_speed(iter/s)": 0.009771
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.7795371498172959,
|
| 431 |
+
"eval_loss": 0.5760576725006104,
|
| 432 |
+
"eval_runtime": 14.2053,
|
| 433 |
+
"eval_samples_per_second": 0.282,
|
| 434 |
+
"eval_steps_per_second": 0.141,
|
| 435 |
+
"step": 200
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.7990255785627284,
|
| 439 |
+
"grad_norm": 0.2861402630805969,
|
| 440 |
+
"learning_rate": 1.0480366524062042e-05,
|
| 441 |
+
"loss": 0.5975393772125244,
|
| 442 |
+
"memory(GiB)": 77.63,
|
| 443 |
+
"step": 205,
|
| 444 |
+
"token_acc": 0.8152735624641129,
|
| 445 |
+
"train_speed(iter/s)": 0.009771
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.8185140073081608,
|
| 449 |
+
"grad_norm": 0.30515122413635254,
|
| 450 |
+
"learning_rate": 8.584303253381847e-06,
|
| 451 |
+
"loss": 0.6246855735778809,
|
| 452 |
+
"memory(GiB)": 77.63,
|
| 453 |
+
"step": 210,
|
| 454 |
+
"token_acc": 0.8046837508398119,
|
| 455 |
+
"train_speed(iter/s)": 0.009789
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.8380024360535931,
|
| 459 |
+
"grad_norm": 0.3188154101371765,
|
| 460 |
+
"learning_rate": 6.861237928494579e-06,
|
| 461 |
+
"loss": 0.5521413803100585,
|
| 462 |
+
"memory(GiB)": 77.63,
|
| 463 |
+
"step": 215,
|
| 464 |
+
"token_acc": 0.8276072283910298,
|
| 465 |
+
"train_speed(iter/s)": 0.009786
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.8574908647990256,
|
| 469 |
+
"grad_norm": 0.3368200361728668,
|
| 470 |
+
"learning_rate": 5.318367983829392e-06,
|
| 471 |
+
"loss": 0.5954047679901123,
|
| 472 |
+
"memory(GiB)": 77.63,
|
| 473 |
+
"step": 220,
|
| 474 |
+
"token_acc": 0.8130686517783292,
|
| 475 |
+
"train_speed(iter/s)": 0.009788
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.876979293544458,
|
| 479 |
+
"grad_norm": 0.3377918601036072,
|
| 480 |
+
"learning_rate": 3.962138157783085e-06,
|
| 481 |
+
"loss": 0.6312428951263428,
|
| 482 |
+
"memory(GiB)": 77.63,
|
| 483 |
+
"step": 225,
|
| 484 |
+
"token_acc": 0.8009544725061238,
|
| 485 |
+
"train_speed(iter/s)": 0.009783
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8964677222898904,
|
| 489 |
+
"grad_norm": 0.34851112961769104,
|
| 490 |
+
"learning_rate": 2.798213572335001e-06,
|
| 491 |
+
"loss": 0.5601920127868653,
|
| 492 |
+
"memory(GiB)": 77.63,
|
| 493 |
+
"step": 230,
|
| 494 |
+
"token_acc": 0.8258236434108527,
|
| 495 |
+
"train_speed(iter/s)": 0.009778
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.9159561510353228,
|
| 499 |
+
"grad_norm": 0.3354482650756836,
|
| 500 |
+
"learning_rate": 1.8314560692059835e-06,
|
| 501 |
+
"loss": 0.5771200656890869,
|
| 502 |
+
"memory(GiB)": 77.63,
|
| 503 |
+
"step": 235,
|
| 504 |
+
"token_acc": 0.8201089968594125,
|
| 505 |
+
"train_speed(iter/s)": 0.009786
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.9354445797807551,
|
| 509 |
+
"grad_norm": 0.29914259910583496,
|
| 510 |
+
"learning_rate": 1.0659039014077944e-06,
|
| 511 |
+
"loss": 0.6100361347198486,
|
| 512 |
+
"memory(GiB)": 77.63,
|
| 513 |
+
"step": 240,
|
| 514 |
+
"token_acc": 0.8085165989634403,
|
| 515 |
+
"train_speed(iter/s)": 0.009798
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.9549330085261876,
|
| 519 |
+
"grad_norm": 0.36409541964530945,
|
| 520 |
+
"learning_rate": 5.047548650136513e-07,
|
| 521 |
+
"loss": 0.6194799423217774,
|
| 522 |
+
"memory(GiB)": 77.63,
|
| 523 |
+
"step": 245,
|
| 524 |
+
"token_acc": 0.807660813392776,
|
| 525 |
+
"train_speed(iter/s)": 0.009801
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.97442143727162,
|
| 529 |
+
"grad_norm": 0.40651243925094604,
|
| 530 |
+
"learning_rate": 1.503529416103988e-07,
|
| 531 |
+
"loss": 0.5671232223510743,
|
| 532 |
+
"memory(GiB)": 77.63,
|
| 533 |
+
"step": 250,
|
| 534 |
+
"token_acc": 0.8216745683871545,
|
| 535 |
+
"train_speed(iter/s)": 0.009809
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.9939098660170523,
|
| 539 |
+
"grad_norm": 0.3596270978450775,
|
| 540 |
+
"learning_rate": 4.178507228136397e-09,
|
| 541 |
+
"loss": 0.6255767822265625,
|
| 542 |
+
"memory(GiB)": 77.63,
|
| 543 |
+
"step": 255,
|
| 544 |
+
"token_acc": 0.8013068013068013,
|
| 545 |
+
"train_speed(iter/s)": 0.009788
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.9978075517661389,
|
| 549 |
+
"eval_loss": 0.5726350545883179,
|
| 550 |
+
"eval_runtime": 14.0103,
|
| 551 |
+
"eval_samples_per_second": 0.286,
|
| 552 |
+
"eval_steps_per_second": 0.143,
|
| 553 |
+
"step": 256
|
| 554 |
+
}
|
| 555 |
+
],
|
| 556 |
+
"logging_steps": 5,
|
| 557 |
+
"max_steps": 256,
|
| 558 |
+
"num_input_tokens_seen": 0,
|
| 559 |
+
"num_train_epochs": 1,
|
| 560 |
+
"save_steps": 50,
|
| 561 |
+
"stateful_callbacks": {
|
| 562 |
+
"TrainerControl": {
|
| 563 |
+
"args": {
|
| 564 |
+
"should_epoch_stop": false,
|
| 565 |
+
"should_evaluate": false,
|
| 566 |
+
"should_log": false,
|
| 567 |
+
"should_save": true,
|
| 568 |
+
"should_training_stop": true
|
| 569 |
+
},
|
| 570 |
+
"attributes": {}
|
| 571 |
+
}
|
| 572 |
+
},
|
| 573 |
+
"total_flos": 1.3517329026916086e+18,
|
| 574 |
+
"train_batch_size": 1,
|
| 575 |
+
"trial_name": null,
|
| 576 |
+
"trial_params": null
|
| 577 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45ea9cbee06f408dfc4621ffd928ab710798716a3b78cc757f3da447b7a31ed0
|
| 3 |
+
size 5816
|