condition-2-zh-20k
#21
by rafaym - opened
tiny-aya-base/condition-2-zh-20k-seed123/adapter_config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": {
|
| 6 |
+
"base_model_class": "Cohere2ForCausalLM",
|
| 7 |
+
"parent_library": "unsloth_compiled_module_cohere2",
|
| 8 |
+
"unsloth_fixed": true
|
| 9 |
+
},
|
| 10 |
+
"base_model_name_or_path": "CohereLabs/tiny-aya-base",
|
| 11 |
+
"bias": "none",
|
| 12 |
+
"corda_config": null,
|
| 13 |
+
"ensure_weight_tying": false,
|
| 14 |
+
"eva_config": null,
|
| 15 |
+
"exclude_modules": null,
|
| 16 |
+
"fan_in_fan_out": false,
|
| 17 |
+
"inference_mode": true,
|
| 18 |
+
"init_lora_weights": true,
|
| 19 |
+
"layer_replication": null,
|
| 20 |
+
"layers_pattern": null,
|
| 21 |
+
"layers_to_transform": null,
|
| 22 |
+
"loftq_config": {},
|
| 23 |
+
"lora_alpha": 32,
|
| 24 |
+
"lora_bias": false,
|
| 25 |
+
"lora_dropout": 0.0,
|
| 26 |
+
"lora_ga_config": null,
|
| 27 |
+
"megatron_config": null,
|
| 28 |
+
"megatron_core": "megatron.core",
|
| 29 |
+
"modules_to_save": null,
|
| 30 |
+
"monteclora_config": null,
|
| 31 |
+
"peft_type": "LORA",
|
| 32 |
+
"peft_version": "0.20.0",
|
| 33 |
+
"qalora_group_size": 16,
|
| 34 |
+
"r": 16,
|
| 35 |
+
"rank_pattern": {},
|
| 36 |
+
"revision": null,
|
| 37 |
+
"target_modules": [
|
| 38 |
+
"o_proj",
|
| 39 |
+
"gate_proj",
|
| 40 |
+
"down_proj",
|
| 41 |
+
"k_proj",
|
| 42 |
+
"q_proj",
|
| 43 |
+
"up_proj",
|
| 44 |
+
"v_proj"
|
| 45 |
+
],
|
| 46 |
+
"target_parameters": null,
|
| 47 |
+
"task_type": "CAUSAL_LM",
|
| 48 |
+
"trainable_token_indices": null,
|
| 49 |
+
"use_bdlora": null,
|
| 50 |
+
"use_dora": false,
|
| 51 |
+
"use_qalora": false,
|
| 52 |
+
"use_rslora": false,
|
| 53 |
+
"velora_config": null
|
| 54 |
+
}
|
tiny-aya-base/condition-2-zh-20k-seed123/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e2ebbc6b22e49160276eaf5311fb59004a9820e8f6f0b1560ee38d85fabb43
|
| 3 |
+
size 120981200
|
tiny-aya-base/condition-2-zh-20k-seed123/training_metrics.json
ADDED
|
@@ -0,0 +1,809 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_config_name": "tiny-aya-base",
|
| 3 |
+
"condition_name": "condition-2-zh-20k",
|
| 4 |
+
"seed": 123,
|
| 5 |
+
"output_name": "condition-2-zh-20k-seed123",
|
| 6 |
+
"train_result": {
|
| 7 |
+
"train_runtime": 14103.4539,
|
| 8 |
+
"train_samples_per_second": 1.276,
|
| 9 |
+
"train_steps_per_second": 0.08,
|
| 10 |
+
"total_flos": 3.140887021564723e+17,
|
| 11 |
+
"train_loss": 1.0681125729878744,
|
| 12 |
+
"epoch": 1.0
|
| 13 |
+
},
|
| 14 |
+
"log_history": [
|
| 15 |
+
{
|
| 16 |
+
"loss": 1.3851,
|
| 17 |
+
"grad_norm": 0.17078915238380432,
|
| 18 |
+
"learning_rate": 3.157894736842105e-05,
|
| 19 |
+
"epoch": 0.008888888888888889,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"loss": 1.3384,
|
| 24 |
+
"grad_norm": 0.15372037887573242,
|
| 25 |
+
"learning_rate": 6.666666666666667e-05,
|
| 26 |
+
"epoch": 0.017777777777777778,
|
| 27 |
+
"step": 20
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"loss": 1.2513,
|
| 31 |
+
"grad_norm": 0.15751078724861145,
|
| 32 |
+
"learning_rate": 0.0001017543859649123,
|
| 33 |
+
"epoch": 0.02666666666666667,
|
| 34 |
+
"step": 30
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"loss": 1.1734,
|
| 38 |
+
"grad_norm": 0.12848994135856628,
|
| 39 |
+
"learning_rate": 0.0001368421052631579,
|
| 40 |
+
"epoch": 0.035555555555555556,
|
| 41 |
+
"step": 40
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"loss": 1.1257,
|
| 45 |
+
"grad_norm": 0.1388208121061325,
|
| 46 |
+
"learning_rate": 0.00017192982456140353,
|
| 47 |
+
"epoch": 0.044444444444444446,
|
| 48 |
+
"step": 50
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"loss": 1.1418,
|
| 52 |
+
"grad_norm": 0.13755622506141663,
|
| 53 |
+
"learning_rate": 0.0001999982694427025,
|
| 54 |
+
"epoch": 0.05333333333333334,
|
| 55 |
+
"step": 60
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"loss": 1.0581,
|
| 59 |
+
"grad_norm": 0.1291951835155487,
|
| 60 |
+
"learning_rate": 0.00019993770622619782,
|
| 61 |
+
"epoch": 0.06222222222222222,
|
| 62 |
+
"step": 70
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"loss": 1.084,
|
| 66 |
+
"grad_norm": 0.1433172971010208,
|
| 67 |
+
"learning_rate": 0.00019979067503207154,
|
| 68 |
+
"epoch": 0.07111111111111111,
|
| 69 |
+
"step": 80
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"loss": 1.0906,
|
| 73 |
+
"grad_norm": 0.13668349385261536,
|
| 74 |
+
"learning_rate": 0.00019955730307447014,
|
| 75 |
+
"epoch": 0.08,
|
| 76 |
+
"step": 90
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"loss": 1.1094,
|
| 80 |
+
"grad_norm": 0.10570124536752701,
|
| 81 |
+
"learning_rate": 0.0001992377922711879,
|
| 82 |
+
"epoch": 0.08888888888888889,
|
| 83 |
+
"step": 100
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"loss": 1.0505,
|
| 87 |
+
"grad_norm": 0.12125972658395767,
|
| 88 |
+
"learning_rate": 0.00019883241906896388,
|
| 89 |
+
"epoch": 0.09777777777777778,
|
| 90 |
+
"step": 110
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"loss": 1.1392,
|
| 94 |
+
"grad_norm": 0.13476838171482086,
|
| 95 |
+
"learning_rate": 0.00019834153420429478,
|
| 96 |
+
"epoch": 0.10666666666666667,
|
| 97 |
+
"step": 120
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"loss": 1.1503,
|
| 101 |
+
"grad_norm": 0.13545557856559753,
|
| 102 |
+
"learning_rate": 0.00019776556239997146,
|
| 103 |
+
"epoch": 0.11555555555555555,
|
| 104 |
+
"step": 130
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"loss": 1.0501,
|
| 108 |
+
"grad_norm": 0.12481781095266342,
|
| 109 |
+
"learning_rate": 0.0001971050019976005,
|
| 110 |
+
"epoch": 0.12444444444444444,
|
| 111 |
+
"step": 140
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"loss": 1.0929,
|
| 115 |
+
"grad_norm": 0.13806591928005219,
|
| 116 |
+
"learning_rate": 0.00019636042452643,
|
| 117 |
+
"epoch": 0.13333333333333333,
|
| 118 |
+
"step": 150
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"loss": 1.0997,
|
| 122 |
+
"grad_norm": 0.12130588293075562,
|
| 123 |
+
"learning_rate": 0.00019553247420885157,
|
| 124 |
+
"epoch": 0.14222222222222222,
|
| 125 |
+
"step": 160
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"loss": 1.0726,
|
| 129 |
+
"grad_norm": 0.11297676712274551,
|
| 130 |
+
"learning_rate": 0.00019462186740300697,
|
| 131 |
+
"epoch": 0.1511111111111111,
|
| 132 |
+
"step": 170
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"loss": 1.0592,
|
| 136 |
+
"grad_norm": 0.12349770218133926,
|
| 137 |
+
"learning_rate": 0.00019362939198298184,
|
| 138 |
+
"epoch": 0.16,
|
| 139 |
+
"step": 180
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"loss": 1.057,
|
| 143 |
+
"grad_norm": 0.13135689496994019,
|
| 144 |
+
"learning_rate": 0.00019255590665712214,
|
| 145 |
+
"epoch": 0.1688888888888889,
|
| 146 |
+
"step": 190
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"loss": 1.0696,
|
| 150 |
+
"grad_norm": 0.11937297880649567,
|
| 151 |
+
"learning_rate": 0.00019140234022506348,
|
| 152 |
+
"epoch": 0.17777777777777778,
|
| 153 |
+
"step": 200
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"loss": 1.014,
|
| 157 |
+
"grad_norm": 0.1261652261018753,
|
| 158 |
+
"learning_rate": 0.00019016969077411647,
|
| 159 |
+
"epoch": 0.18666666666666668,
|
| 160 |
+
"step": 210
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"loss": 1.049,
|
| 164 |
+
"grad_norm": 0.11738768965005875,
|
| 165 |
+
"learning_rate": 0.0001888590248157027,
|
| 166 |
+
"epoch": 0.19555555555555557,
|
| 167 |
+
"step": 220
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"loss": 1.1171,
|
| 171 |
+
"grad_norm": 0.13960260152816772,
|
| 172 |
+
"learning_rate": 0.00018747147636258917,
|
| 173 |
+
"epoch": 0.20444444444444446,
|
| 174 |
+
"step": 230
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"loss": 1.018,
|
| 178 |
+
"grad_norm": 0.12128771841526031,
|
| 179 |
+
"learning_rate": 0.00018600824594771907,
|
| 180 |
+
"epoch": 0.21333333333333335,
|
| 181 |
+
"step": 240
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"loss": 1.0909,
|
| 185 |
+
"grad_norm": 0.12346600741147995,
|
| 186 |
+
"learning_rate": 0.0001844705995854882,
|
| 187 |
+
"epoch": 0.2222222222222222,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"loss": 1.0709,
|
| 192 |
+
"grad_norm": 0.12290473282337189,
|
| 193 |
+
"learning_rate": 0.00018285986767636566,
|
| 194 |
+
"epoch": 0.2311111111111111,
|
| 195 |
+
"step": 260
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"loss": 1.0941,
|
| 199 |
+
"grad_norm": 0.1199881061911583,
|
| 200 |
+
"learning_rate": 0.00018117744385580625,
|
| 201 |
+
"epoch": 0.24,
|
| 202 |
+
"step": 270
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"loss": 1.0963,
|
| 206 |
+
"grad_norm": 0.13761970400810242,
|
| 207 |
+
"learning_rate": 0.0001794247837884511,
|
| 208 |
+
"epoch": 0.24888888888888888,
|
| 209 |
+
"step": 280
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"loss": 1.0695,
|
| 213 |
+
"grad_norm": 0.13161271810531616,
|
| 214 |
+
"learning_rate": 0.0001776034039086592,
|
| 215 |
+
"epoch": 0.2577777777777778,
|
| 216 |
+
"step": 290
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"loss": 1.0983,
|
| 220 |
+
"grad_norm": 0.10759685933589935,
|
| 221 |
+
"learning_rate": 0.00017571488010846003,
|
| 222 |
+
"epoch": 0.26666666666666666,
|
| 223 |
+
"step": 300
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"loss": 1.0212,
|
| 227 |
+
"grad_norm": 0.14229480922222137,
|
| 228 |
+
"learning_rate": 0.00017376084637406222,
|
| 229 |
+
"epoch": 0.27555555555555555,
|
| 230 |
+
"step": 310
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"loss": 1.0666,
|
| 234 |
+
"grad_norm": 0.133957177400589,
|
| 235 |
+
"learning_rate": 0.000171742993372098,
|
| 236 |
+
"epoch": 0.28444444444444444,
|
| 237 |
+
"step": 320
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"loss": 1.0221,
|
| 241 |
+
"grad_norm": 0.101350799202919,
|
| 242 |
+
"learning_rate": 0.0001696630669868267,
|
| 243 |
+
"epoch": 0.29333333333333333,
|
| 244 |
+
"step": 330
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"loss": 1.0475,
|
| 248 |
+
"grad_norm": 0.10641703009605408,
|
| 249 |
+
"learning_rate": 0.00016752286680956306,
|
| 250 |
+
"epoch": 0.3022222222222222,
|
| 251 |
+
"step": 340
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"loss": 1.0546,
|
| 255 |
+
"grad_norm": 0.12402649968862534,
|
| 256 |
+
"learning_rate": 0.00016532424458163693,
|
| 257 |
+
"epoch": 0.3111111111111111,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"loss": 1.0432,
|
| 262 |
+
"grad_norm": 0.10325026512145996,
|
| 263 |
+
"learning_rate": 0.0001630691025922321,
|
| 264 |
+
"epoch": 0.32,
|
| 265 |
+
"step": 360
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"loss": 1.0808,
|
| 269 |
+
"grad_norm": 0.12462154030799866,
|
| 270 |
+
"learning_rate": 0.0001607593920324899,
|
| 271 |
+
"epoch": 0.3288888888888889,
|
| 272 |
+
"step": 370
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"loss": 1.0732,
|
| 276 |
+
"grad_norm": 0.13195562362670898,
|
| 277 |
+
"learning_rate": 0.00015839711130730203,
|
| 278 |
+
"epoch": 0.3377777777777778,
|
| 279 |
+
"step": 380
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"loss": 1.0432,
|
| 283 |
+
"grad_norm": 0.11475151032209396,
|
| 284 |
+
"learning_rate": 0.00015598430430625334,
|
| 285 |
+
"epoch": 0.3466666666666667,
|
| 286 |
+
"step": 390
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"loss": 1.0842,
|
| 290 |
+
"grad_norm": 0.12205471098423004,
|
| 291 |
+
"learning_rate": 0.00015352305863520991,
|
| 292 |
+
"epoch": 0.35555555555555557,
|
| 293 |
+
"step": 400
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"loss": 1.0525,
|
| 297 |
+
"grad_norm": 0.13711963593959808,
|
| 298 |
+
"learning_rate": 0.00015101550381008377,
|
| 299 |
+
"epoch": 0.36444444444444446,
|
| 300 |
+
"step": 410
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"loss": 1.0075,
|
| 304 |
+
"grad_norm": 0.10483007878065109,
|
| 305 |
+
"learning_rate": 0.00014846380941433522,
|
| 306 |
+
"epoch": 0.37333333333333335,
|
| 307 |
+
"step": 420
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"loss": 1.1111,
|
| 311 |
+
"grad_norm": 0.10803163796663284,
|
| 312 |
+
"learning_rate": 0.00014587018322180905,
|
| 313 |
+
"epoch": 0.38222222222222224,
|
| 314 |
+
"step": 430
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"loss": 1.0544,
|
| 318 |
+
"grad_norm": 0.12365908175706863,
|
| 319 |
+
"learning_rate": 0.00014323686928652697,
|
| 320 |
+
"epoch": 0.39111111111111113,
|
| 321 |
+
"step": 440
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"loss": 1.0502,
|
| 325 |
+
"grad_norm": 0.10243521630764008,
|
| 326 |
+
"learning_rate": 0.00014056614600108997,
|
| 327 |
+
"epoch": 0.4,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"loss": 1.0197,
|
| 332 |
+
"grad_norm": 0.12014426290988922,
|
| 333 |
+
"learning_rate": 0.00013786032412537035,
|
| 334 |
+
"epoch": 0.4088888888888889,
|
| 335 |
+
"step": 460
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"loss": 1.0664,
|
| 339 |
+
"grad_norm": 0.1101444885134697,
|
| 340 |
+
"learning_rate": 0.00013512174478719894,
|
| 341 |
+
"epoch": 0.4177777777777778,
|
| 342 |
+
"step": 470
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"loss": 1.065,
|
| 346 |
+
"grad_norm": 0.10545314848423004,
|
| 347 |
+
"learning_rate": 0.00013235277745677747,
|
| 348 |
+
"epoch": 0.4266666666666667,
|
| 349 |
+
"step": 480
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"loss": 1.0724,
|
| 353 |
+
"grad_norm": 0.13452774286270142,
|
| 354 |
+
"learning_rate": 0.00012955581789656843,
|
| 355 |
+
"epoch": 0.43555555555555553,
|
| 356 |
+
"step": 490
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"loss": 1.1233,
|
| 360 |
+
"grad_norm": 0.11292257905006409,
|
| 361 |
+
"learning_rate": 0.00012673328608843636,
|
| 362 |
+
"epoch": 0.4444444444444444,
|
| 363 |
+
"step": 500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"loss": 1.1052,
|
| 367 |
+
"grad_norm": 0.12459282577037811,
|
| 368 |
+
"learning_rate": 0.00012388762413983445,
|
| 369 |
+
"epoch": 0.4533333333333333,
|
| 370 |
+
"step": 510
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"loss": 1.05,
|
| 374 |
+
"grad_norm": 0.14430229365825653,
|
| 375 |
+
"learning_rate": 0.00012102129417084714,
|
| 376 |
+
"epoch": 0.4622222222222222,
|
| 377 |
+
"step": 520
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"loss": 1.036,
|
| 381 |
+
"grad_norm": 0.10886573791503906,
|
| 382 |
+
"learning_rate": 0.00011813677618391759,
|
| 383 |
+
"epoch": 0.4711111111111111,
|
| 384 |
+
"step": 530
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"loss": 0.9994,
|
| 388 |
+
"grad_norm": 0.10799901187419891,
|
| 389 |
+
"learning_rate": 0.00011523656591810337,
|
| 390 |
+
"epoch": 0.48,
|
| 391 |
+
"step": 540
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"loss": 1.0771,
|
| 395 |
+
"grad_norm": 0.13042695820331573,
|
| 396 |
+
"learning_rate": 0.00011232317268971585,
|
| 397 |
+
"epoch": 0.4888888888888889,
|
| 398 |
+
"step": 550
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"loss": 1.0494,
|
| 402 |
+
"grad_norm": 0.1067989394068718,
|
| 403 |
+
"learning_rate": 0.00010939911722121306,
|
| 404 |
+
"epoch": 0.49777777777777776,
|
| 405 |
+
"step": 560
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"loss": 1.0622,
|
| 409 |
+
"grad_norm": 0.1132044568657875,
|
| 410 |
+
"learning_rate": 0.00010646692946022285,
|
| 411 |
+
"epoch": 0.5066666666666667,
|
| 412 |
+
"step": 570
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"loss": 1.0864,
|
| 416 |
+
"grad_norm": 0.12146525830030441,
|
| 417 |
+
"learning_rate": 0.00010352914639058526,
|
| 418 |
+
"epoch": 0.5155555555555555,
|
| 419 |
+
"step": 580
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"loss": 1.0359,
|
| 423 |
+
"grad_norm": 0.10813307017087936,
|
| 424 |
+
"learning_rate": 0.00010058830983730622,
|
| 425 |
+
"epoch": 0.5244444444444445,
|
| 426 |
+
"step": 590
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"loss": 1.095,
|
| 430 |
+
"grad_norm": 0.11109290271997452,
|
| 431 |
+
"learning_rate": 9.764696426732303e-05,
|
| 432 |
+
"epoch": 0.5333333333333333,
|
| 433 |
+
"step": 600
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"loss": 1.0217,
|
| 437 |
+
"grad_norm": 0.1386152058839798,
|
| 438 |
+
"learning_rate": 9.470765458798368e-05,
|
| 439 |
+
"epoch": 0.5422222222222223,
|
| 440 |
+
"step": 610
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"loss": 1.0327,
|
| 444 |
+
"grad_norm": 0.1302192509174347,
|
| 445 |
+
"learning_rate": 9.177292394514555e-05,
|
| 446 |
+
"epoch": 0.5511111111111111,
|
| 447 |
+
"step": 620
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"loss": 1.0478,
|
| 451 |
+
"grad_norm": 0.12203508615493774,
|
| 452 |
+
"learning_rate": 8.884531152279756e-05,
|
| 453 |
+
"epoch": 0.56,
|
| 454 |
+
"step": 630
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"loss": 1.0588,
|
| 458 |
+
"grad_norm": 0.10961955040693283,
|
| 459 |
+
"learning_rate": 8.592735034611097e-05,
|
| 460 |
+
"epoch": 0.5688888888888889,
|
| 461 |
+
"step": 640
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"loss": 1.0704,
|
| 465 |
+
"grad_norm": 0.1224837377667427,
|
| 466 |
+
"learning_rate": 8.302156508981815e-05,
|
| 467 |
+
"epoch": 0.5777777777777777,
|
| 468 |
+
"step": 650
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"loss": 1.0604,
|
| 472 |
+
"grad_norm": 0.13877613842487335,
|
| 473 |
+
"learning_rate": 8.013046989381691e-05,
|
| 474 |
+
"epoch": 0.5866666666666667,
|
| 475 |
+
"step": 660
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"loss": 1.0581,
|
| 479 |
+
"grad_norm": 0.11927918344736099,
|
| 480 |
+
"learning_rate": 7.725656618788937e-05,
|
| 481 |
+
"epoch": 0.5955555555555555,
|
| 482 |
+
"step": 670
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"loss": 1.0587,
|
| 486 |
+
"grad_norm": 0.11579909920692444,
|
| 487 |
+
"learning_rate": 7.4402340527418e-05,
|
| 488 |
+
"epoch": 0.6044444444444445,
|
| 489 |
+
"step": 680
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"loss": 1.0124,
|
| 493 |
+
"grad_norm": 0.11753726750612259,
|
| 494 |
+
"learning_rate": 7.157026244197132e-05,
|
| 495 |
+
"epoch": 0.6133333333333333,
|
| 496 |
+
"step": 690
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"loss": 1.0636,
|
| 500 |
+
"grad_norm": 0.11829619854688644,
|
| 501 |
+
"learning_rate": 6.87627822986206e-05,
|
| 502 |
+
"epoch": 0.6222222222222222,
|
| 503 |
+
"step": 700
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"loss": 1.0418,
|
| 507 |
+
"grad_norm": 0.11234968155622482,
|
| 508 |
+
"learning_rate": 6.598232918183632e-05,
|
| 509 |
+
"epoch": 0.6311111111111111,
|
| 510 |
+
"step": 710
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"loss": 1.0525,
|
| 514 |
+
"grad_norm": 0.13825488090515137,
|
| 515 |
+
"learning_rate": 6.323130879179875e-05,
|
| 516 |
+
"epoch": 0.64,
|
| 517 |
+
"step": 720
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"loss": 1.0824,
|
| 521 |
+
"grad_norm": 0.12125255912542343,
|
| 522 |
+
"learning_rate": 6.051210136294089e-05,
|
| 523 |
+
"epoch": 0.6488888888888888,
|
| 524 |
+
"step": 730
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"loss": 1.0357,
|
| 528 |
+
"grad_norm": 0.11368292570114136,
|
| 529 |
+
"learning_rate": 5.7827059604525234e-05,
|
| 530 |
+
"epoch": 0.6577777777777778,
|
| 531 |
+
"step": 740
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"loss": 1.0809,
|
| 535 |
+
"grad_norm": 0.11407827585935593,
|
| 536 |
+
"learning_rate": 5.517850666503547e-05,
|
| 537 |
+
"epoch": 0.6666666666666666,
|
| 538 |
+
"step": 750
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"loss": 1.0148,
|
| 542 |
+
"grad_norm": 0.10886642336845398,
|
| 543 |
+
"learning_rate": 5.2568734122144756e-05,
|
| 544 |
+
"epoch": 0.6755555555555556,
|
| 545 |
+
"step": 760
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"loss": 1.0438,
|
| 549 |
+
"grad_norm": 0.13094542920589447,
|
| 550 |
+
"learning_rate": 5.000000000000002e-05,
|
| 551 |
+
"epoch": 0.6844444444444444,
|
| 552 |
+
"step": 770
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"loss": 1.0285,
|
| 556 |
+
"grad_norm": 0.11394693702459335,
|
| 557 |
+
"learning_rate": 4.747452681553674e-05,
|
| 558 |
+
"epoch": 0.6933333333333334,
|
| 559 |
+
"step": 780
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"loss": 1.0833,
|
| 563 |
+
"grad_norm": 0.1083562970161438,
|
| 564 |
+
"learning_rate": 4.4994499655515865e-05,
|
| 565 |
+
"epoch": 0.7022222222222222,
|
| 566 |
+
"step": 790
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"loss": 1.0622,
|
| 570 |
+
"grad_norm": 0.11369741708040237,
|
| 571 |
+
"learning_rate": 4.256206428594587e-05,
|
| 572 |
+
"epoch": 0.7111111111111111,
|
| 573 |
+
"step": 800
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"loss": 1.0251,
|
| 577 |
+
"grad_norm": 0.13660773634910583,
|
| 578 |
+
"learning_rate": 4.017932529552543e-05,
|
| 579 |
+
"epoch": 0.72,
|
| 580 |
+
"step": 810
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"loss": 1.0154,
|
| 584 |
+
"grad_norm": 0.10507376492023468,
|
| 585 |
+
"learning_rate": 3.784834427471408e-05,
|
| 586 |
+
"epoch": 0.7288888888888889,
|
| 587 |
+
"step": 820
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"loss": 1.0853,
|
| 591 |
+
"grad_norm": 0.11955918371677399,
|
| 592 |
+
"learning_rate": 3.557113803200537e-05,
|
| 593 |
+
"epoch": 0.7377777777777778,
|
| 594 |
+
"step": 830
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"loss": 1.1013,
|
| 598 |
+
"grad_norm": 0.12790556252002716,
|
| 599 |
+
"learning_rate": 3.3349676848946345e-05,
|
| 600 |
+
"epoch": 0.7466666666666667,
|
| 601 |
+
"step": 840
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"loss": 1.0148,
|
| 605 |
+
"grad_norm": 0.11285852640867233,
|
| 606 |
+
"learning_rate": 3.118588277541312e-05,
|
| 607 |
+
"epoch": 0.7555555555555555,
|
| 608 |
+
"step": 850
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"loss": 1.0874,
|
| 612 |
+
"grad_norm": 0.11276240646839142,
|
| 613 |
+
"learning_rate": 2.9081627966617096e-05,
|
| 614 |
+
"epoch": 0.7644444444444445,
|
| 615 |
+
"step": 860
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"loss": 1.0573,
|
| 619 |
+
"grad_norm": 0.11087917536497116,
|
| 620 |
+
"learning_rate": 2.7038733063281174e-05,
|
| 621 |
+
"epoch": 0.7733333333333333,
|
| 622 |
+
"step": 870
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"loss": 0.9906,
|
| 626 |
+
"grad_norm": 0.12071377038955688,
|
| 627 |
+
"learning_rate": 2.5058965616387498e-05,
|
| 628 |
+
"epoch": 0.7822222222222223,
|
| 629 |
+
"step": 880
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"loss": 1.0645,
|
| 633 |
+
"grad_norm": 0.10874035209417343,
|
| 634 |
+
"learning_rate": 2.3144038557858916e-05,
|
| 635 |
+
"epoch": 0.7911111111111111,
|
| 636 |
+
"step": 890
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"loss": 1.022,
|
| 640 |
+
"grad_norm": 0.11100213974714279,
|
| 641 |
+
"learning_rate": 2.1295608718498284e-05,
|
| 642 |
+
"epoch": 0.8,
|
| 643 |
+
"step": 900
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"loss": 1.0458,
|
| 647 |
+
"grad_norm": 0.10863592475652695,
|
| 648 |
+
"learning_rate": 1.9515275394467446e-05,
|
| 649 |
+
"epoch": 0.8088888888888889,
|
| 650 |
+
"step": 910
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"loss": 1.0784,
|
| 654 |
+
"grad_norm": 0.11265172809362411,
|
| 655 |
+
"learning_rate": 1.7804578963545994e-05,
|
| 656 |
+
"epoch": 0.8177777777777778,
|
| 657 |
+
"step": 920
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"loss": 1.0393,
|
| 661 |
+
"grad_norm": 0.11236457526683807,
|
| 662 |
+
"learning_rate": 1.6164999552367765e-05,
|
| 663 |
+
"epoch": 0.8266666666666667,
|
| 664 |
+
"step": 930
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"loss": 1.0318,
|
| 668 |
+
"grad_norm": 0.11550150066614151,
|
| 669 |
+
"learning_rate": 1.4597955755787373e-05,
|
| 670 |
+
"epoch": 0.8355555555555556,
|
| 671 |
+
"step": 940
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"loss": 1.0925,
|
| 675 |
+
"grad_norm": 0.12814396619796753,
|
| 676 |
+
"learning_rate": 1.3104803409485356e-05,
|
| 677 |
+
"epoch": 0.8444444444444444,
|
| 678 |
+
"step": 950
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"loss": 1.0248,
|
| 682 |
+
"grad_norm": 0.12026702612638474,
|
| 683 |
+
"learning_rate": 1.1686834416873815e-05,
|
| 684 |
+
"epoch": 0.8533333333333334,
|
| 685 |
+
"step": 960
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"loss": 1.0438,
|
| 689 |
+
"grad_norm": 0.11860033124685287,
|
| 690 |
+
"learning_rate": 1.0345275631317163e-05,
|
| 691 |
+
"epoch": 0.8622222222222222,
|
| 692 |
+
"step": 970
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"loss": 1.0678,
|
| 696 |
+
"grad_norm": 0.11572486162185669,
|
| 697 |
+
"learning_rate": 9.081287794635774e-06,
|
| 698 |
+
"epoch": 0.8711111111111111,
|
| 699 |
+
"step": 980
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"loss": 1.0762,
|
| 703 |
+
"grad_norm": 0.12252067774534225,
|
| 704 |
+
"learning_rate": 7.895964532810317e-06,
|
| 705 |
+
"epoch": 0.88,
|
| 706 |
+
"step": 990
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"loss": 1.0365,
|
| 710 |
+
"grad_norm": 0.12507659196853638,
|
| 711 |
+
"learning_rate": 6.7903314097560454e-06,
|
| 712 |
+
"epoch": 0.8888888888888888,
|
| 713 |
+
"step": 1000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"loss": 1.043,
|
| 717 |
+
"grad_norm": 0.12245221436023712,
|
| 718 |
+
"learning_rate": 5.765345039985648e-06,
|
| 719 |
+
"epoch": 0.8977777777777778,
|
| 720 |
+
"step": 1010
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"loss": 1.0449,
|
| 724 |
+
"grad_norm": 0.12368922680616379,
|
| 725 |
+
"learning_rate": 4.821892260928451e-06,
|
| 726 |
+
"epoch": 0.9066666666666666,
|
| 727 |
+
"step": 1020
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"loss": 1.0025,
|
| 731 |
+
"grad_norm": 0.0971086323261261,
|
| 732 |
+
"learning_rate": 3.960789365622075e-06,
|
| 733 |
+
"epoch": 0.9155555555555556,
|
| 734 |
+
"step": 1030
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"loss": 1.0795,
|
| 738 |
+
"grad_norm": 0.10536965727806091,
|
| 739 |
+
"learning_rate": 3.1827813964403484e-06,
|
| 740 |
+
"epoch": 0.9244444444444444,
|
| 741 |
+
"step": 1040
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"loss": 1.0739,
|
| 745 |
+
"grad_norm": 0.11942451447248459,
|
| 746 |
+
"learning_rate": 2.4885415004686665e-06,
|
| 747 |
+
"epoch": 0.9333333333333333,
|
| 748 |
+
"step": 1050
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"loss": 0.9879,
|
| 752 |
+
"grad_norm": 0.11502082645893097,
|
| 753 |
+
"learning_rate": 1.8786703470845547e-06,
|
| 754 |
+
"epoch": 0.9422222222222222,
|
| 755 |
+
"step": 1060
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"loss": 0.9971,
|
| 759 |
+
"grad_norm": 0.11270863562822342,
|
| 760 |
+
"learning_rate": 1.3536956082472074e-06,
|
| 761 |
+
"epoch": 0.9511111111111111,
|
| 762 |
+
"step": 1070
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"loss": 1.1131,
|
| 766 |
+
"grad_norm": 0.1083143875002861,
|
| 767 |
+
"learning_rate": 9.140715019458457e-07,
|
| 768 |
+
"epoch": 0.96,
|
| 769 |
+
"step": 1080
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"loss": 1.0692,
|
| 773 |
+
"grad_norm": 0.12779924273490906,
|
| 774 |
+
"learning_rate": 5.60178399201805e-07,
|
| 775 |
+
"epoch": 0.9688888888888889,
|
| 776 |
+
"step": 1090
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"loss": 1.0959,
|
| 780 |
+
"grad_norm": 0.12438718229532242,
|
| 781 |
+
"learning_rate": 2.923224949643477e-07,
|
| 782 |
+
"epoch": 0.9777777777777777,
|
| 783 |
+
"step": 1100
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"loss": 1.0366,
|
| 787 |
+
"grad_norm": 0.11403634399175644,
|
| 788 |
+
"learning_rate": 1.1073554318509205e-07,
|
| 789 |
+
"epoch": 0.9866666666666667,
|
| 790 |
+
"step": 1110
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"loss": 1.0828,
|
| 794 |
+
"grad_norm": 0.12552717328071594,
|
| 795 |
+
"learning_rate": 1.5574656300143542e-08,
|
| 796 |
+
"epoch": 0.9955555555555555,
|
| 797 |
+
"step": 1120
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"train_runtime": 14103.4539,
|
| 801 |
+
"train_samples_per_second": 1.276,
|
| 802 |
+
"train_steps_per_second": 0.08,
|
| 803 |
+
"total_flos": 3.140887021564723e+17,
|
| 804 |
+
"train_loss": 1.0681125729878744,
|
| 805 |
+
"epoch": 1.0,
|
| 806 |
+
"step": 1125
|
| 807 |
+
}
|
| 808 |
+
]
|
| 809 |
+
}
|