Upload fine-tuned SigLIP2 model for ai-vs-hum
Browse files- all_results.json +13 -0
- checkpoint-160000/config.json +47 -0
- checkpoint-160000/model.safetensors +3 -0
- checkpoint-160000/optimizer.pt +3 -0
- checkpoint-160000/rng_state.pth +3 -0
- checkpoint-160000/scheduler.pt +3 -0
- checkpoint-160000/trainer_state.json +2310 -0
- checkpoint-160000/training_args.bin +3 -0
- checkpoint-200000/config.json +47 -0
- checkpoint-200000/model.safetensors +3 -0
- checkpoint-200000/optimizer.pt +3 -0
- checkpoint-200000/rng_state.pth +3 -0
- checkpoint-200000/scheduler.pt +3 -0
- checkpoint-200000/trainer_state.json +2879 -0
- checkpoint-200000/training_args.bin +3 -0
- config.json +47 -0
- eval_results.json +8 -0
- model.safetensors +3 -0
- runs/Nov16_11-57-36_blacc/events.out.tfevents.1763283456.blacc +3 -0
- runs/Nov16_11-57-36_blacc/events.out.tfevents.1763289551.blacc +3 -0
- train_results.json +8 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.98965,
|
| 4 |
+
"eval_loss": 0.10565917938947678,
|
| 5 |
+
"eval_runtime": 70.587,
|
| 6 |
+
"eval_samples_per_second": 283.338,
|
| 7 |
+
"eval_steps_per_second": 70.835,
|
| 8 |
+
"total_flos": 3.35021983285248e+19,
|
| 9 |
+
"train_loss": 0.12970913526404707,
|
| 10 |
+
"train_runtime": 6024.3105,
|
| 11 |
+
"train_samples_per_second": 66.398,
|
| 12 |
+
"train_steps_per_second": 33.199
|
| 13 |
+
}
|
checkpoint-160000/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"dtype": "float32",
|
| 6 |
+
"id2label": {
|
| 7 |
+
"0": "ai",
|
| 8 |
+
"1": "hum"
|
| 9 |
+
},
|
| 10 |
+
"initializer_factor": 1.0,
|
| 11 |
+
"label2id": {
|
| 12 |
+
"ai": 0,
|
| 13 |
+
"hum": 1
|
| 14 |
+
},
|
| 15 |
+
"model_type": "siglip",
|
| 16 |
+
"problem_type": "single_label_classification",
|
| 17 |
+
"text_config": {
|
| 18 |
+
"attention_dropout": 0.0,
|
| 19 |
+
"dtype": "float32",
|
| 20 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"layer_norm_eps": 1e-06,
|
| 24 |
+
"max_position_embeddings": 64,
|
| 25 |
+
"model_type": "siglip_text_model",
|
| 26 |
+
"num_attention_heads": 12,
|
| 27 |
+
"num_hidden_layers": 12,
|
| 28 |
+
"projection_size": 768,
|
| 29 |
+
"vocab_size": 256000
|
| 30 |
+
},
|
| 31 |
+
"transformers_version": "5.0.0.dev0",
|
| 32 |
+
"use_cache": false,
|
| 33 |
+
"vision_config": {
|
| 34 |
+
"attention_dropout": 0.0,
|
| 35 |
+
"dtype": "float32",
|
| 36 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 37 |
+
"hidden_size": 768,
|
| 38 |
+
"image_size": 224,
|
| 39 |
+
"intermediate_size": 3072,
|
| 40 |
+
"layer_norm_eps": 1e-06,
|
| 41 |
+
"model_type": "siglip_vision_model",
|
| 42 |
+
"num_attention_heads": 12,
|
| 43 |
+
"num_channels": 3,
|
| 44 |
+
"num_hidden_layers": 12,
|
| 45 |
+
"patch_size": 16
|
| 46 |
+
}
|
| 47 |
+
}
|
checkpoint-160000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4fef8677fb140086f0d7a58555255722851534a79d8ef0b67b036bd2f7f3058
|
| 3 |
+
size 371567992
|
checkpoint-160000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d36d9f8aa7df2a2871a449f71ae3586dbb61176afe1ba81118592c07538b21cc
|
| 3 |
+
size 686558987
|
checkpoint-160000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ef8c5c3269dc6616dd312ba898bf2532055b6f070e164aa5569264c73937d57
|
| 3 |
+
size 14645
|
checkpoint-160000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9486643d2e1909246fdf5a43b7f96c7956c1730b43609fee18b77fa724848559
|
| 3 |
+
size 1465
|
checkpoint-160000/trainer_state.json
ADDED
|
@@ -0,0 +1,2310 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 120000,
|
| 3 |
+
"best_metric": 0.9856,
|
| 4 |
+
"best_model_checkpoint": "ai-vs-hum/checkpoint-120000",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 160000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0125,
|
| 14 |
+
"grad_norm": 57.61789321899414,
|
| 15 |
+
"learning_rate": 1.2475e-07,
|
| 16 |
+
"loss": 1.0591,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.025,
|
| 21 |
+
"grad_norm": 69.49858093261719,
|
| 22 |
+
"learning_rate": 2.4975e-07,
|
| 23 |
+
"loss": 0.5393,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0375,
|
| 28 |
+
"grad_norm": 1023.806640625,
|
| 29 |
+
"learning_rate": 3.7475000000000004e-07,
|
| 30 |
+
"loss": 0.4548,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05,
|
| 35 |
+
"grad_norm": 777.0330810546875,
|
| 36 |
+
"learning_rate": 4.997500000000001e-07,
|
| 37 |
+
"loss": 0.6269,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0625,
|
| 42 |
+
"grad_norm": 0.08837065100669861,
|
| 43 |
+
"learning_rate": 6.2475e-07,
|
| 44 |
+
"loss": 0.5631,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.075,
|
| 49 |
+
"grad_norm": 0.024585120379924774,
|
| 50 |
+
"learning_rate": 7.497500000000001e-07,
|
| 51 |
+
"loss": 0.5938,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0875,
|
| 56 |
+
"grad_norm": 0.012241708114743233,
|
| 57 |
+
"learning_rate": 8.747500000000001e-07,
|
| 58 |
+
"loss": 0.5345,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1,
|
| 63 |
+
"grad_norm": 0.37598320841789246,
|
| 64 |
+
"learning_rate": 9.9975e-07,
|
| 65 |
+
"loss": 0.4773,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1125,
|
| 70 |
+
"grad_norm": 0.11799755692481995,
|
| 71 |
+
"learning_rate": 1.1247500000000001e-06,
|
| 72 |
+
"loss": 0.4258,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.125,
|
| 77 |
+
"grad_norm": 0.034392450004816055,
|
| 78 |
+
"learning_rate": 1.2497500000000002e-06,
|
| 79 |
+
"loss": 0.578,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1375,
|
| 84 |
+
"grad_norm": 775.758056640625,
|
| 85 |
+
"learning_rate": 1.3747500000000003e-06,
|
| 86 |
+
"loss": 0.5597,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.15,
|
| 91 |
+
"grad_norm": 0.01906261406838894,
|
| 92 |
+
"learning_rate": 1.4997500000000001e-06,
|
| 93 |
+
"loss": 0.4832,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1625,
|
| 98 |
+
"grad_norm": 0.016003865748643875,
|
| 99 |
+
"learning_rate": 1.6247500000000002e-06,
|
| 100 |
+
"loss": 0.4621,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.175,
|
| 105 |
+
"grad_norm": 0.052252043038606644,
|
| 106 |
+
"learning_rate": 1.74975e-06,
|
| 107 |
+
"loss": 0.4056,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1875,
|
| 112 |
+
"grad_norm": 0.008017883636057377,
|
| 113 |
+
"learning_rate": 1.8747500000000001e-06,
|
| 114 |
+
"loss": 0.4268,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2,
|
| 119 |
+
"grad_norm": 0.04491295665502548,
|
| 120 |
+
"learning_rate": 1.99975e-06,
|
| 121 |
+
"loss": 0.4946,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2125,
|
| 126 |
+
"grad_norm": 0.24870972335338593,
|
| 127 |
+
"learning_rate": 2.1247500000000003e-06,
|
| 128 |
+
"loss": 0.3448,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.225,
|
| 133 |
+
"grad_norm": 0.39745479822158813,
|
| 134 |
+
"learning_rate": 2.2497500000000004e-06,
|
| 135 |
+
"loss": 0.5246,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2375,
|
| 140 |
+
"grad_norm": 0.1810966432094574,
|
| 141 |
+
"learning_rate": 2.37475e-06,
|
| 142 |
+
"loss": 0.3799,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.25,
|
| 147 |
+
"grad_norm": 0.03164760395884514,
|
| 148 |
+
"learning_rate": 2.49975e-06,
|
| 149 |
+
"loss": 0.3887,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.2625,
|
| 154 |
+
"grad_norm": 22.039918899536133,
|
| 155 |
+
"learning_rate": 2.62475e-06,
|
| 156 |
+
"loss": 0.3956,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.275,
|
| 161 |
+
"grad_norm": 0.06578505039215088,
|
| 162 |
+
"learning_rate": 2.7497500000000007e-06,
|
| 163 |
+
"loss": 0.4013,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2875,
|
| 168 |
+
"grad_norm": 0.016265813261270523,
|
| 169 |
+
"learning_rate": 2.87475e-06,
|
| 170 |
+
"loss": 0.3442,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3,
|
| 175 |
+
"grad_norm": 0.016749408096075058,
|
| 176 |
+
"learning_rate": 2.99975e-06,
|
| 177 |
+
"loss": 0.404,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3125,
|
| 182 |
+
"grad_norm": 0.0033249990083277225,
|
| 183 |
+
"learning_rate": 3.1247500000000004e-06,
|
| 184 |
+
"loss": 0.3088,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.325,
|
| 189 |
+
"grad_norm": 0.3450980484485626,
|
| 190 |
+
"learning_rate": 3.2497500000000005e-06,
|
| 191 |
+
"loss": 0.4177,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3375,
|
| 196 |
+
"grad_norm": 0.01602231152355671,
|
| 197 |
+
"learning_rate": 3.3747500000000006e-06,
|
| 198 |
+
"loss": 0.351,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.35,
|
| 203 |
+
"grad_norm": 0.00211876118555665,
|
| 204 |
+
"learning_rate": 3.4997500000000002e-06,
|
| 205 |
+
"loss": 0.3653,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3625,
|
| 210 |
+
"grad_norm": 0.023136630654335022,
|
| 211 |
+
"learning_rate": 3.6247500000000003e-06,
|
| 212 |
+
"loss": 0.3439,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.375,
|
| 217 |
+
"grad_norm": 0.002042749198153615,
|
| 218 |
+
"learning_rate": 3.7497500000000004e-06,
|
| 219 |
+
"loss": 0.3052,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.3875,
|
| 224 |
+
"grad_norm": 0.023861056193709373,
|
| 225 |
+
"learning_rate": 3.8747500000000005e-06,
|
| 226 |
+
"loss": 0.3397,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4,
|
| 231 |
+
"grad_norm": 0.13855504989624023,
|
| 232 |
+
"learning_rate": 3.999750000000001e-06,
|
| 233 |
+
"loss": 0.2901,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.4125,
|
| 238 |
+
"grad_norm": 0.05347450450062752,
|
| 239 |
+
"learning_rate": 4.12475e-06,
|
| 240 |
+
"loss": 0.3356,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.425,
|
| 245 |
+
"grad_norm": 186.85972595214844,
|
| 246 |
+
"learning_rate": 4.24975e-06,
|
| 247 |
+
"loss": 0.4037,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.4375,
|
| 252 |
+
"grad_norm": 0.009446755982935429,
|
| 253 |
+
"learning_rate": 4.374750000000001e-06,
|
| 254 |
+
"loss": 0.3582,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.45,
|
| 259 |
+
"grad_norm": 0.02892524003982544,
|
| 260 |
+
"learning_rate": 4.49975e-06,
|
| 261 |
+
"loss": 0.2834,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4625,
|
| 266 |
+
"grad_norm": 0.024663692340254784,
|
| 267 |
+
"learning_rate": 4.624750000000001e-06,
|
| 268 |
+
"loss": 0.3806,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.475,
|
| 273 |
+
"grad_norm": 0.17350807785987854,
|
| 274 |
+
"learning_rate": 4.7497500000000006e-06,
|
| 275 |
+
"loss": 0.3418,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.4875,
|
| 280 |
+
"grad_norm": 0.02292953059077263,
|
| 281 |
+
"learning_rate": 4.87475e-06,
|
| 282 |
+
"loss": 0.3369,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.5,
|
| 287 |
+
"grad_norm": 0.014667114242911339,
|
| 288 |
+
"learning_rate": 4.999750000000001e-06,
|
| 289 |
+
"loss": 0.3386,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5125,
|
| 294 |
+
"grad_norm": 0.012080417945981026,
|
| 295 |
+
"learning_rate": 4.986138888888889e-06,
|
| 296 |
+
"loss": 0.3399,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.525,
|
| 301 |
+
"grad_norm": 0.012398626655340195,
|
| 302 |
+
"learning_rate": 4.9722500000000005e-06,
|
| 303 |
+
"loss": 0.3269,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5375,
|
| 308 |
+
"grad_norm": 0.7642549276351929,
|
| 309 |
+
"learning_rate": 4.958361111111111e-06,
|
| 310 |
+
"loss": 0.2879,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.55,
|
| 315 |
+
"grad_norm": 0.09296104311943054,
|
| 316 |
+
"learning_rate": 4.9444722222222226e-06,
|
| 317 |
+
"loss": 0.3181,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5625,
|
| 322 |
+
"grad_norm": 0.00763101689517498,
|
| 323 |
+
"learning_rate": 4.930583333333334e-06,
|
| 324 |
+
"loss": 0.2816,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.575,
|
| 329 |
+
"grad_norm": 0.01172051951289177,
|
| 330 |
+
"learning_rate": 4.916694444444445e-06,
|
| 331 |
+
"loss": 0.3259,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5875,
|
| 336 |
+
"grad_norm": 0.015954310074448586,
|
| 337 |
+
"learning_rate": 4.902805555555556e-06,
|
| 338 |
+
"loss": 0.2336,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.6,
|
| 343 |
+
"grad_norm": 20.73999786376953,
|
| 344 |
+
"learning_rate": 4.8889166666666674e-06,
|
| 345 |
+
"loss": 0.3255,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.6125,
|
| 350 |
+
"grad_norm": 0.01183092501014471,
|
| 351 |
+
"learning_rate": 4.875027777777778e-06,
|
| 352 |
+
"loss": 0.2953,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.625,
|
| 357 |
+
"grad_norm": 368.5484619140625,
|
| 358 |
+
"learning_rate": 4.8611388888888895e-06,
|
| 359 |
+
"loss": 0.2733,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6375,
|
| 364 |
+
"grad_norm": 0.023890795186161995,
|
| 365 |
+
"learning_rate": 4.847250000000001e-06,
|
| 366 |
+
"loss": 0.2929,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.65,
|
| 371 |
+
"grad_norm": 0.009666858240962029,
|
| 372 |
+
"learning_rate": 4.8333611111111115e-06,
|
| 373 |
+
"loss": 0.2289,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6625,
|
| 378 |
+
"grad_norm": 11.71799373626709,
|
| 379 |
+
"learning_rate": 4.819472222222223e-06,
|
| 380 |
+
"loss": 0.2396,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.675,
|
| 385 |
+
"grad_norm": 173.8921356201172,
|
| 386 |
+
"learning_rate": 4.8055833333333335e-06,
|
| 387 |
+
"loss": 0.2684,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6875,
|
| 392 |
+
"grad_norm": 0.0010307455668225884,
|
| 393 |
+
"learning_rate": 4.791694444444445e-06,
|
| 394 |
+
"loss": 0.2543,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.7,
|
| 399 |
+
"grad_norm": 0.5270601511001587,
|
| 400 |
+
"learning_rate": 4.7778055555555555e-06,
|
| 401 |
+
"loss": 0.2652,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7125,
|
| 406 |
+
"grad_norm": 0.031072473153471947,
|
| 407 |
+
"learning_rate": 4.763916666666667e-06,
|
| 408 |
+
"loss": 0.291,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.725,
|
| 413 |
+
"grad_norm": 0.015740619972348213,
|
| 414 |
+
"learning_rate": 4.750027777777778e-06,
|
| 415 |
+
"loss": 0.2381,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7375,
|
| 420 |
+
"grad_norm": 0.013210597448050976,
|
| 421 |
+
"learning_rate": 4.736138888888889e-06,
|
| 422 |
+
"loss": 0.239,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.75,
|
| 427 |
+
"grad_norm": 0.00735945301130414,
|
| 428 |
+
"learning_rate": 4.72225e-06,
|
| 429 |
+
"loss": 0.1984,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7625,
|
| 434 |
+
"grad_norm": 0.011138584464788437,
|
| 435 |
+
"learning_rate": 4.708361111111112e-06,
|
| 436 |
+
"loss": 0.2401,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.775,
|
| 441 |
+
"grad_norm": 3.5947213172912598,
|
| 442 |
+
"learning_rate": 4.694472222222222e-06,
|
| 443 |
+
"loss": 0.2375,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7875,
|
| 448 |
+
"grad_norm": 0.03575507923960686,
|
| 449 |
+
"learning_rate": 4.680583333333334e-06,
|
| 450 |
+
"loss": 0.264,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.8,
|
| 455 |
+
"grad_norm": 0.011910141445696354,
|
| 456 |
+
"learning_rate": 4.666694444444445e-06,
|
| 457 |
+
"loss": 0.2367,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.8125,
|
| 462 |
+
"grad_norm": 0.015016453340649605,
|
| 463 |
+
"learning_rate": 4.652805555555556e-06,
|
| 464 |
+
"loss": 0.2427,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.825,
|
| 469 |
+
"grad_norm": 129.52890014648438,
|
| 470 |
+
"learning_rate": 4.6389166666666665e-06,
|
| 471 |
+
"loss": 0.2208,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8375,
|
| 476 |
+
"grad_norm": 0.03964228183031082,
|
| 477 |
+
"learning_rate": 4.625027777777778e-06,
|
| 478 |
+
"loss": 0.1863,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.85,
|
| 483 |
+
"grad_norm": 0.06889275461435318,
|
| 484 |
+
"learning_rate": 4.611138888888889e-06,
|
| 485 |
+
"loss": 0.2526,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8625,
|
| 490 |
+
"grad_norm": 0.009284596890211105,
|
| 491 |
+
"learning_rate": 4.59725e-06,
|
| 492 |
+
"loss": 0.2281,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.875,
|
| 497 |
+
"grad_norm": 0.02173932082951069,
|
| 498 |
+
"learning_rate": 4.583361111111111e-06,
|
| 499 |
+
"loss": 0.2357,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8875,
|
| 504 |
+
"grad_norm": 0.2086459994316101,
|
| 505 |
+
"learning_rate": 4.569472222222223e-06,
|
| 506 |
+
"loss": 0.1764,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.9,
|
| 511 |
+
"grad_norm": 0.3403266668319702,
|
| 512 |
+
"learning_rate": 4.555583333333334e-06,
|
| 513 |
+
"loss": 0.2075,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.9125,
|
| 518 |
+
"grad_norm": 0.010942379012703896,
|
| 519 |
+
"learning_rate": 4.541694444444445e-06,
|
| 520 |
+
"loss": 0.2245,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.925,
|
| 525 |
+
"grad_norm": 8.935309597291052e-05,
|
| 526 |
+
"learning_rate": 4.527805555555556e-06,
|
| 527 |
+
"loss": 0.1583,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.9375,
|
| 532 |
+
"grad_norm": 0.0007859256002120674,
|
| 533 |
+
"learning_rate": 4.513916666666668e-06,
|
| 534 |
+
"loss": 0.2016,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.95,
|
| 539 |
+
"grad_norm": 9.582052734913304e-05,
|
| 540 |
+
"learning_rate": 4.500027777777778e-06,
|
| 541 |
+
"loss": 0.2117,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9625,
|
| 546 |
+
"grad_norm": 801.5213012695312,
|
| 547 |
+
"learning_rate": 4.486138888888889e-06,
|
| 548 |
+
"loss": 0.1858,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.975,
|
| 553 |
+
"grad_norm": 0.00019180079107172787,
|
| 554 |
+
"learning_rate": 4.47225e-06,
|
| 555 |
+
"loss": 0.2273,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9875,
|
| 560 |
+
"grad_norm": 0.00854406412690878,
|
| 561 |
+
"learning_rate": 4.458361111111112e-06,
|
| 562 |
+
"loss": 0.2419,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.0,
|
| 567 |
+
"grad_norm": 0.5351490378379822,
|
| 568 |
+
"learning_rate": 4.444472222222222e-06,
|
| 569 |
+
"loss": 0.1964,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.0,
|
| 574 |
+
"eval_accuracy": 0.91205,
|
| 575 |
+
"eval_loss": 0.5335493087768555,
|
| 576 |
+
"eval_runtime": 72.3355,
|
| 577 |
+
"eval_samples_per_second": 276.489,
|
| 578 |
+
"eval_steps_per_second": 69.122,
|
| 579 |
+
"step": 40000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.0125,
|
| 583 |
+
"grad_norm": 0.010720465332269669,
|
| 584 |
+
"learning_rate": 4.430583333333334e-06,
|
| 585 |
+
"loss": 0.1343,
|
| 586 |
+
"step": 40500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.025,
|
| 590 |
+
"grad_norm": 0.001053882297128439,
|
| 591 |
+
"learning_rate": 4.416694444444445e-06,
|
| 592 |
+
"loss": 0.2317,
|
| 593 |
+
"step": 41000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.0375,
|
| 597 |
+
"grad_norm": 0.08478885889053345,
|
| 598 |
+
"learning_rate": 4.402805555555556e-06,
|
| 599 |
+
"loss": 0.1304,
|
| 600 |
+
"step": 41500
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.05,
|
| 604 |
+
"grad_norm": 0.007641951087862253,
|
| 605 |
+
"learning_rate": 4.388916666666667e-06,
|
| 606 |
+
"loss": 0.1912,
|
| 607 |
+
"step": 42000
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 1.0625,
|
| 611 |
+
"grad_norm": 0.0008450501482002437,
|
| 612 |
+
"learning_rate": 4.375027777777779e-06,
|
| 613 |
+
"loss": 0.1816,
|
| 614 |
+
"step": 42500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.075,
|
| 618 |
+
"grad_norm": 0.00660599023103714,
|
| 619 |
+
"learning_rate": 4.361138888888889e-06,
|
| 620 |
+
"loss": 0.1872,
|
| 621 |
+
"step": 43000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.0875,
|
| 625 |
+
"grad_norm": 0.02802414819598198,
|
| 626 |
+
"learning_rate": 4.34725e-06,
|
| 627 |
+
"loss": 0.1797,
|
| 628 |
+
"step": 43500
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.1,
|
| 632 |
+
"grad_norm": 0.04336298257112503,
|
| 633 |
+
"learning_rate": 4.333361111111111e-06,
|
| 634 |
+
"loss": 0.1473,
|
| 635 |
+
"step": 44000
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.1125,
|
| 639 |
+
"grad_norm": 0.0009568212553858757,
|
| 640 |
+
"learning_rate": 4.319472222222223e-06,
|
| 641 |
+
"loss": 0.1572,
|
| 642 |
+
"step": 44500
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.125,
|
| 646 |
+
"grad_norm": 0.001199259888380766,
|
| 647 |
+
"learning_rate": 4.305583333333333e-06,
|
| 648 |
+
"loss": 0.1874,
|
| 649 |
+
"step": 45000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.1375,
|
| 653 |
+
"grad_norm": 0.008190172724425793,
|
| 654 |
+
"learning_rate": 4.291694444444445e-06,
|
| 655 |
+
"loss": 0.2571,
|
| 656 |
+
"step": 45500
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.15,
|
| 660 |
+
"grad_norm": 0.018085084855556488,
|
| 661 |
+
"learning_rate": 4.277805555555556e-06,
|
| 662 |
+
"loss": 0.2197,
|
| 663 |
+
"step": 46000
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.1625,
|
| 667 |
+
"grad_norm": 0.00019737149705179036,
|
| 668 |
+
"learning_rate": 4.263916666666667e-06,
|
| 669 |
+
"loss": 0.1498,
|
| 670 |
+
"step": 46500
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.175,
|
| 674 |
+
"grad_norm": 0.0007391768740490079,
|
| 675 |
+
"learning_rate": 4.250027777777778e-06,
|
| 676 |
+
"loss": 0.1599,
|
| 677 |
+
"step": 47000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.1875,
|
| 681 |
+
"grad_norm": 0.00027270030113868415,
|
| 682 |
+
"learning_rate": 4.2361388888888895e-06,
|
| 683 |
+
"loss": 0.1637,
|
| 684 |
+
"step": 47500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.2,
|
| 688 |
+
"grad_norm": 0.007702063303440809,
|
| 689 |
+
"learning_rate": 4.22225e-06,
|
| 690 |
+
"loss": 0.2234,
|
| 691 |
+
"step": 48000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.2125,
|
| 695 |
+
"grad_norm": 76.7024917602539,
|
| 696 |
+
"learning_rate": 4.2083611111111115e-06,
|
| 697 |
+
"loss": 0.2164,
|
| 698 |
+
"step": 48500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.225,
|
| 702 |
+
"grad_norm": 0.005129904951900244,
|
| 703 |
+
"learning_rate": 4.194472222222223e-06,
|
| 704 |
+
"loss": 0.0964,
|
| 705 |
+
"step": 49000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.2375,
|
| 709 |
+
"grad_norm": 0.005428163334727287,
|
| 710 |
+
"learning_rate": 4.1805833333333336e-06,
|
| 711 |
+
"loss": 0.2027,
|
| 712 |
+
"step": 49500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.25,
|
| 716 |
+
"grad_norm": 0.0023394206073135138,
|
| 717 |
+
"learning_rate": 4.166694444444445e-06,
|
| 718 |
+
"loss": 0.1597,
|
| 719 |
+
"step": 50000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.2625,
|
| 723 |
+
"grad_norm": 3.80388087251049e-06,
|
| 724 |
+
"learning_rate": 4.1528055555555556e-06,
|
| 725 |
+
"loss": 0.1041,
|
| 726 |
+
"step": 50500
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.275,
|
| 730 |
+
"grad_norm": 0.10043651610612869,
|
| 731 |
+
"learning_rate": 4.138916666666667e-06,
|
| 732 |
+
"loss": 0.1689,
|
| 733 |
+
"step": 51000
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 1.2875,
|
| 737 |
+
"grad_norm": 0.8555727601051331,
|
| 738 |
+
"learning_rate": 4.1250277777777784e-06,
|
| 739 |
+
"loss": 0.1732,
|
| 740 |
+
"step": 51500
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.3,
|
| 744 |
+
"grad_norm": 0.036154795438051224,
|
| 745 |
+
"learning_rate": 4.111138888888889e-06,
|
| 746 |
+
"loss": 0.1895,
|
| 747 |
+
"step": 52000
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.3125,
|
| 751 |
+
"grad_norm": 0.12704850733280182,
|
| 752 |
+
"learning_rate": 4.0972500000000005e-06,
|
| 753 |
+
"loss": 0.1476,
|
| 754 |
+
"step": 52500
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.325,
|
| 758 |
+
"grad_norm": 0.037445101886987686,
|
| 759 |
+
"learning_rate": 4.083361111111112e-06,
|
| 760 |
+
"loss": 0.1434,
|
| 761 |
+
"step": 53000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.3375,
|
| 765 |
+
"grad_norm": 1.4244892554415856e-05,
|
| 766 |
+
"learning_rate": 4.0694722222222225e-06,
|
| 767 |
+
"loss": 0.1934,
|
| 768 |
+
"step": 53500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.35,
|
| 772 |
+
"grad_norm": 0.05814299359917641,
|
| 773 |
+
"learning_rate": 4.055583333333334e-06,
|
| 774 |
+
"loss": 0.1674,
|
| 775 |
+
"step": 54000
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 1.3625,
|
| 779 |
+
"grad_norm": 120.2337417602539,
|
| 780 |
+
"learning_rate": 4.041694444444445e-06,
|
| 781 |
+
"loss": 0.155,
|
| 782 |
+
"step": 54500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.375,
|
| 786 |
+
"grad_norm": 0.012538187205791473,
|
| 787 |
+
"learning_rate": 4.027805555555556e-06,
|
| 788 |
+
"loss": 0.1196,
|
| 789 |
+
"step": 55000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1.3875,
|
| 793 |
+
"grad_norm": 0.01885361410677433,
|
| 794 |
+
"learning_rate": 4.0139166666666665e-06,
|
| 795 |
+
"loss": 0.1773,
|
| 796 |
+
"step": 55500
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.4,
|
| 800 |
+
"grad_norm": 0.0020437000785022974,
|
| 801 |
+
"learning_rate": 4.000027777777778e-06,
|
| 802 |
+
"loss": 0.2195,
|
| 803 |
+
"step": 56000
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.4125,
|
| 807 |
+
"grad_norm": 0.0043403045274317265,
|
| 808 |
+
"learning_rate": 3.986138888888889e-06,
|
| 809 |
+
"loss": 0.1506,
|
| 810 |
+
"step": 56500
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.425,
|
| 814 |
+
"grad_norm": 0.029600977897644043,
|
| 815 |
+
"learning_rate": 3.97225e-06,
|
| 816 |
+
"loss": 0.1905,
|
| 817 |
+
"step": 57000
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 1.4375,
|
| 821 |
+
"grad_norm": 0.0036520950961858034,
|
| 822 |
+
"learning_rate": 3.958361111111111e-06,
|
| 823 |
+
"loss": 0.1869,
|
| 824 |
+
"step": 57500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.45,
|
| 828 |
+
"grad_norm": 0.0004638723039533943,
|
| 829 |
+
"learning_rate": 3.944472222222223e-06,
|
| 830 |
+
"loss": 0.1575,
|
| 831 |
+
"step": 58000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 1.4625,
|
| 835 |
+
"grad_norm": 0.00012263878306839615,
|
| 836 |
+
"learning_rate": 3.930583333333333e-06,
|
| 837 |
+
"loss": 0.1624,
|
| 838 |
+
"step": 58500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.475,
|
| 842 |
+
"grad_norm": 132.2960205078125,
|
| 843 |
+
"learning_rate": 3.916694444444445e-06,
|
| 844 |
+
"loss": 0.1547,
|
| 845 |
+
"step": 59000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.4875,
|
| 849 |
+
"grad_norm": 0.002475355053320527,
|
| 850 |
+
"learning_rate": 3.902805555555556e-06,
|
| 851 |
+
"loss": 0.1517,
|
| 852 |
+
"step": 59500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.5,
|
| 856 |
+
"grad_norm": 0.396793931722641,
|
| 857 |
+
"learning_rate": 3.888916666666667e-06,
|
| 858 |
+
"loss": 0.1749,
|
| 859 |
+
"step": 60000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 1.5125,
|
| 863 |
+
"grad_norm": 0.0002843483234755695,
|
| 864 |
+
"learning_rate": 3.8750277777777774e-06,
|
| 865 |
+
"loss": 0.1465,
|
| 866 |
+
"step": 60500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.525,
|
| 870 |
+
"grad_norm": 0.013508698903024197,
|
| 871 |
+
"learning_rate": 3.861138888888889e-06,
|
| 872 |
+
"loss": 0.1794,
|
| 873 |
+
"step": 61000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 1.5375,
|
| 877 |
+
"grad_norm": 0.3459691107273102,
|
| 878 |
+
"learning_rate": 3.84725e-06,
|
| 879 |
+
"loss": 0.1431,
|
| 880 |
+
"step": 61500
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 1.55,
|
| 884 |
+
"grad_norm": 0.0022266607265919447,
|
| 885 |
+
"learning_rate": 3.833361111111111e-06,
|
| 886 |
+
"loss": 0.1725,
|
| 887 |
+
"step": 62000
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 1.5625,
|
| 891 |
+
"grad_norm": 0.015393017791211605,
|
| 892 |
+
"learning_rate": 3.819472222222222e-06,
|
| 893 |
+
"loss": 0.1659,
|
| 894 |
+
"step": 62500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.575,
|
| 898 |
+
"grad_norm": 0.0006895596161484718,
|
| 899 |
+
"learning_rate": 3.8055833333333338e-06,
|
| 900 |
+
"loss": 0.104,
|
| 901 |
+
"step": 63000
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 1.5875,
|
| 905 |
+
"grad_norm": 99.34886932373047,
|
| 906 |
+
"learning_rate": 3.7916944444444448e-06,
|
| 907 |
+
"loss": 0.1676,
|
| 908 |
+
"step": 63500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.6,
|
| 912 |
+
"grad_norm": 0.0647934079170227,
|
| 913 |
+
"learning_rate": 3.7778055555555558e-06,
|
| 914 |
+
"loss": 0.1576,
|
| 915 |
+
"step": 64000
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 1.6125,
|
| 919 |
+
"grad_norm": 0.0019897823221981525,
|
| 920 |
+
"learning_rate": 3.763916666666667e-06,
|
| 921 |
+
"loss": 0.1288,
|
| 922 |
+
"step": 64500
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.625,
|
| 926 |
+
"grad_norm": 0.0011682398617267609,
|
| 927 |
+
"learning_rate": 3.7500277777777782e-06,
|
| 928 |
+
"loss": 0.1508,
|
| 929 |
+
"step": 65000
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.6375,
|
| 933 |
+
"grad_norm": 0.006074314471334219,
|
| 934 |
+
"learning_rate": 3.7361388888888892e-06,
|
| 935 |
+
"loss": 0.1355,
|
| 936 |
+
"step": 65500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.65,
|
| 940 |
+
"grad_norm": 0.0016712377546355128,
|
| 941 |
+
"learning_rate": 3.7222500000000002e-06,
|
| 942 |
+
"loss": 0.1515,
|
| 943 |
+
"step": 66000
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 1.6625,
|
| 947 |
+
"grad_norm": 0.022902656346559525,
|
| 948 |
+
"learning_rate": 3.7083611111111112e-06,
|
| 949 |
+
"loss": 0.1502,
|
| 950 |
+
"step": 66500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.675,
|
| 954 |
+
"grad_norm": 1.1224092304473743e-05,
|
| 955 |
+
"learning_rate": 3.6944722222222223e-06,
|
| 956 |
+
"loss": 0.1649,
|
| 957 |
+
"step": 67000
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.6875,
|
| 961 |
+
"grad_norm": 0.00015980980242602527,
|
| 962 |
+
"learning_rate": 3.6805833333333337e-06,
|
| 963 |
+
"loss": 0.184,
|
| 964 |
+
"step": 67500
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.7,
|
| 968 |
+
"grad_norm": 0.008271472528576851,
|
| 969 |
+
"learning_rate": 3.6666944444444447e-06,
|
| 970 |
+
"loss": 0.1036,
|
| 971 |
+
"step": 68000
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.7125,
|
| 975 |
+
"grad_norm": 0.003849911969155073,
|
| 976 |
+
"learning_rate": 3.6528055555555557e-06,
|
| 977 |
+
"loss": 0.1225,
|
| 978 |
+
"step": 68500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.725,
|
| 982 |
+
"grad_norm": 0.00014347155229188502,
|
| 983 |
+
"learning_rate": 3.638916666666667e-06,
|
| 984 |
+
"loss": 0.1213,
|
| 985 |
+
"step": 69000
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.7375,
|
| 989 |
+
"grad_norm": 0.0031831359956413507,
|
| 990 |
+
"learning_rate": 3.625027777777778e-06,
|
| 991 |
+
"loss": 0.1658,
|
| 992 |
+
"step": 69500
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.75,
|
| 996 |
+
"grad_norm": 0.00018772439216263592,
|
| 997 |
+
"learning_rate": 3.611138888888889e-06,
|
| 998 |
+
"loss": 0.1193,
|
| 999 |
+
"step": 70000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.7625,
|
| 1003 |
+
"grad_norm": 0.0017813814338296652,
|
| 1004 |
+
"learning_rate": 3.5972500000000006e-06,
|
| 1005 |
+
"loss": 0.1326,
|
| 1006 |
+
"step": 70500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.775,
|
| 1010 |
+
"grad_norm": 0.006025127600878477,
|
| 1011 |
+
"learning_rate": 3.5833611111111116e-06,
|
| 1012 |
+
"loss": 0.0978,
|
| 1013 |
+
"step": 71000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.7875,
|
| 1017 |
+
"grad_norm": 0.10651105642318726,
|
| 1018 |
+
"learning_rate": 3.5694722222222226e-06,
|
| 1019 |
+
"loss": 0.163,
|
| 1020 |
+
"step": 71500
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.8,
|
| 1024 |
+
"grad_norm": 0.0005099499830976129,
|
| 1025 |
+
"learning_rate": 3.555583333333333e-06,
|
| 1026 |
+
"loss": 0.1218,
|
| 1027 |
+
"step": 72000
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.8125,
|
| 1031 |
+
"grad_norm": 0.013583320192992687,
|
| 1032 |
+
"learning_rate": 3.5416944444444446e-06,
|
| 1033 |
+
"loss": 0.0974,
|
| 1034 |
+
"step": 72500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.825,
|
| 1038 |
+
"grad_norm": 0.004824354313313961,
|
| 1039 |
+
"learning_rate": 3.5278055555555556e-06,
|
| 1040 |
+
"loss": 0.1885,
|
| 1041 |
+
"step": 73000
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.8375,
|
| 1045 |
+
"grad_norm": 0.017673935741186142,
|
| 1046 |
+
"learning_rate": 3.513916666666667e-06,
|
| 1047 |
+
"loss": 0.1381,
|
| 1048 |
+
"step": 73500
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.85,
|
| 1052 |
+
"grad_norm": 3.5670293527800823e-06,
|
| 1053 |
+
"learning_rate": 3.500027777777778e-06,
|
| 1054 |
+
"loss": 0.1369,
|
| 1055 |
+
"step": 74000
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.8625,
|
| 1059 |
+
"grad_norm": 0.03615639731287956,
|
| 1060 |
+
"learning_rate": 3.486138888888889e-06,
|
| 1061 |
+
"loss": 0.1453,
|
| 1062 |
+
"step": 74500
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.875,
|
| 1066 |
+
"grad_norm": 0.00019953282026108354,
|
| 1067 |
+
"learning_rate": 3.4722500000000005e-06,
|
| 1068 |
+
"loss": 0.117,
|
| 1069 |
+
"step": 75000
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 1.8875,
|
| 1073 |
+
"grad_norm": 9.114891952322068e-08,
|
| 1074 |
+
"learning_rate": 3.4583611111111115e-06,
|
| 1075 |
+
"loss": 0.1227,
|
| 1076 |
+
"step": 75500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.9,
|
| 1080 |
+
"grad_norm": 0.0070320055820047855,
|
| 1081 |
+
"learning_rate": 3.4444722222222225e-06,
|
| 1082 |
+
"loss": 0.0892,
|
| 1083 |
+
"step": 76000
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.9125,
|
| 1087 |
+
"grad_norm": 2.0117835447308607e-05,
|
| 1088 |
+
"learning_rate": 3.430583333333334e-06,
|
| 1089 |
+
"loss": 0.1272,
|
| 1090 |
+
"step": 76500
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.925,
|
| 1094 |
+
"grad_norm": 1.597428672539536e-05,
|
| 1095 |
+
"learning_rate": 3.416694444444445e-06,
|
| 1096 |
+
"loss": 0.1555,
|
| 1097 |
+
"step": 77000
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.9375,
|
| 1101 |
+
"grad_norm": 0.03601458668708801,
|
| 1102 |
+
"learning_rate": 3.4028055555555556e-06,
|
| 1103 |
+
"loss": 0.1518,
|
| 1104 |
+
"step": 77500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.95,
|
| 1108 |
+
"grad_norm": 68.35214233398438,
|
| 1109 |
+
"learning_rate": 3.3889166666666666e-06,
|
| 1110 |
+
"loss": 0.1481,
|
| 1111 |
+
"step": 78000
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.9625,
|
| 1115 |
+
"grad_norm": 0.03208393603563309,
|
| 1116 |
+
"learning_rate": 3.375027777777778e-06,
|
| 1117 |
+
"loss": 0.1816,
|
| 1118 |
+
"step": 78500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.975,
|
| 1122 |
+
"grad_norm": 0.04399827867746353,
|
| 1123 |
+
"learning_rate": 3.361138888888889e-06,
|
| 1124 |
+
"loss": 0.149,
|
| 1125 |
+
"step": 79000
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.9875,
|
| 1129 |
+
"grad_norm": 0.007236841134727001,
|
| 1130 |
+
"learning_rate": 3.34725e-06,
|
| 1131 |
+
"loss": 0.1421,
|
| 1132 |
+
"step": 79500
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.0,
|
| 1136 |
+
"grad_norm": 475.6263427734375,
|
| 1137 |
+
"learning_rate": 3.3333611111111114e-06,
|
| 1138 |
+
"loss": 0.1498,
|
| 1139 |
+
"step": 80000
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.0,
|
| 1143 |
+
"eval_accuracy": 0.98015,
|
| 1144 |
+
"eval_loss": 0.11060585081577301,
|
| 1145 |
+
"eval_runtime": 71.1478,
|
| 1146 |
+
"eval_samples_per_second": 281.105,
|
| 1147 |
+
"eval_steps_per_second": 70.276,
|
| 1148 |
+
"step": 80000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.0125,
|
| 1152 |
+
"grad_norm": 8.577239896112587e-06,
|
| 1153 |
+
"learning_rate": 3.3194722222222225e-06,
|
| 1154 |
+
"loss": 0.0677,
|
| 1155 |
+
"step": 80500
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 2.025,
|
| 1159 |
+
"grad_norm": 2.4226756067946553e-07,
|
| 1160 |
+
"learning_rate": 3.305583333333334e-06,
|
| 1161 |
+
"loss": 0.0873,
|
| 1162 |
+
"step": 81000
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 2.0375,
|
| 1166 |
+
"grad_norm": 0.0019143010722473264,
|
| 1167 |
+
"learning_rate": 3.291694444444445e-06,
|
| 1168 |
+
"loss": 0.1247,
|
| 1169 |
+
"step": 81500
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 2.05,
|
| 1173 |
+
"grad_norm": 0.00012104820052627474,
|
| 1174 |
+
"learning_rate": 3.277805555555556e-06,
|
| 1175 |
+
"loss": 0.0585,
|
| 1176 |
+
"step": 82000
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.0625,
|
| 1180 |
+
"grad_norm": 0.02980652078986168,
|
| 1181 |
+
"learning_rate": 3.2639166666666673e-06,
|
| 1182 |
+
"loss": 0.0722,
|
| 1183 |
+
"step": 82500
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 2.075,
|
| 1187 |
+
"grad_norm": 0.00023673578107263893,
|
| 1188 |
+
"learning_rate": 3.250027777777778e-06,
|
| 1189 |
+
"loss": 0.1211,
|
| 1190 |
+
"step": 83000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.0875,
|
| 1194 |
+
"grad_norm": 0.020454034209251404,
|
| 1195 |
+
"learning_rate": 3.236138888888889e-06,
|
| 1196 |
+
"loss": 0.0758,
|
| 1197 |
+
"step": 83500
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 2.1,
|
| 1201 |
+
"grad_norm": 2.698595835681772e-06,
|
| 1202 |
+
"learning_rate": 3.22225e-06,
|
| 1203 |
+
"loss": 0.0527,
|
| 1204 |
+
"step": 84000
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 2.1125,
|
| 1208 |
+
"grad_norm": 0.0003798000980168581,
|
| 1209 |
+
"learning_rate": 3.2083611111111114e-06,
|
| 1210 |
+
"loss": 0.1335,
|
| 1211 |
+
"step": 84500
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 2.125,
|
| 1215 |
+
"grad_norm": 0.0007345640915445983,
|
| 1216 |
+
"learning_rate": 3.1944722222222224e-06,
|
| 1217 |
+
"loss": 0.1113,
|
| 1218 |
+
"step": 85000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 2.1375,
|
| 1222 |
+
"grad_norm": 0.09974172711372375,
|
| 1223 |
+
"learning_rate": 3.1805833333333334e-06,
|
| 1224 |
+
"loss": 0.0601,
|
| 1225 |
+
"step": 85500
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 2.15,
|
| 1229 |
+
"grad_norm": 0.018733147531747818,
|
| 1230 |
+
"learning_rate": 3.166694444444445e-06,
|
| 1231 |
+
"loss": 0.1287,
|
| 1232 |
+
"step": 86000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.1625,
|
| 1236 |
+
"grad_norm": 0.00047316361451521516,
|
| 1237 |
+
"learning_rate": 3.152805555555556e-06,
|
| 1238 |
+
"loss": 0.0931,
|
| 1239 |
+
"step": 86500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 2.175,
|
| 1243 |
+
"grad_norm": 0.00016918570327106863,
|
| 1244 |
+
"learning_rate": 3.1389166666666673e-06,
|
| 1245 |
+
"loss": 0.0967,
|
| 1246 |
+
"step": 87000
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 2.1875,
|
| 1250 |
+
"grad_norm": 1.8574954083305784e-05,
|
| 1251 |
+
"learning_rate": 3.1250277777777783e-06,
|
| 1252 |
+
"loss": 0.0994,
|
| 1253 |
+
"step": 87500
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 2.2,
|
| 1257 |
+
"grad_norm": 9.292660251958296e-05,
|
| 1258 |
+
"learning_rate": 3.1111388888888893e-06,
|
| 1259 |
+
"loss": 0.0941,
|
| 1260 |
+
"step": 88000
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 2.2125,
|
| 1264 |
+
"grad_norm": 2.387504309808719e-06,
|
| 1265 |
+
"learning_rate": 3.09725e-06,
|
| 1266 |
+
"loss": 0.0806,
|
| 1267 |
+
"step": 88500
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 2.225,
|
| 1271 |
+
"grad_norm": 0.011588960886001587,
|
| 1272 |
+
"learning_rate": 3.0833611111111113e-06,
|
| 1273 |
+
"loss": 0.1497,
|
| 1274 |
+
"step": 89000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.2375,
|
| 1278 |
+
"grad_norm": 0.007856378331780434,
|
| 1279 |
+
"learning_rate": 3.0694722222222223e-06,
|
| 1280 |
+
"loss": 0.1076,
|
| 1281 |
+
"step": 89500
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 2.25,
|
| 1285 |
+
"grad_norm": 0.014046729542315006,
|
| 1286 |
+
"learning_rate": 3.0555833333333333e-06,
|
| 1287 |
+
"loss": 0.0754,
|
| 1288 |
+
"step": 90000
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 2.2625,
|
| 1292 |
+
"grad_norm": 0.13570299744606018,
|
| 1293 |
+
"learning_rate": 3.0416944444444447e-06,
|
| 1294 |
+
"loss": 0.0647,
|
| 1295 |
+
"step": 90500
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 2.275,
|
| 1299 |
+
"grad_norm": 1.2789449783667806e-06,
|
| 1300 |
+
"learning_rate": 3.0278055555555558e-06,
|
| 1301 |
+
"loss": 0.0433,
|
| 1302 |
+
"step": 91000
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 2.2875,
|
| 1306 |
+
"grad_norm": 0.00010895993909798563,
|
| 1307 |
+
"learning_rate": 3.0139166666666668e-06,
|
| 1308 |
+
"loss": 0.0798,
|
| 1309 |
+
"step": 91500
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 2.3,
|
| 1313 |
+
"grad_norm": 0.0030339015647768974,
|
| 1314 |
+
"learning_rate": 3.000027777777778e-06,
|
| 1315 |
+
"loss": 0.1252,
|
| 1316 |
+
"step": 92000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.3125,
|
| 1320 |
+
"grad_norm": 0.12522538006305695,
|
| 1321 |
+
"learning_rate": 2.986138888888889e-06,
|
| 1322 |
+
"loss": 0.0911,
|
| 1323 |
+
"step": 92500
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 2.325,
|
| 1327 |
+
"grad_norm": 0.00940883718430996,
|
| 1328 |
+
"learning_rate": 2.9722500000000002e-06,
|
| 1329 |
+
"loss": 0.0999,
|
| 1330 |
+
"step": 93000
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 2.3375,
|
| 1334 |
+
"grad_norm": 0.05087483301758766,
|
| 1335 |
+
"learning_rate": 2.9583611111111116e-06,
|
| 1336 |
+
"loss": 0.1069,
|
| 1337 |
+
"step": 93500
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 2.35,
|
| 1341 |
+
"grad_norm": 0.0046332115307450294,
|
| 1342 |
+
"learning_rate": 2.9444722222222227e-06,
|
| 1343 |
+
"loss": 0.0814,
|
| 1344 |
+
"step": 94000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 2.3625,
|
| 1348 |
+
"grad_norm": 4.976536729373038e-05,
|
| 1349 |
+
"learning_rate": 2.9305833333333332e-06,
|
| 1350 |
+
"loss": 0.0749,
|
| 1351 |
+
"step": 94500
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 2.375,
|
| 1355 |
+
"grad_norm": 0.13259227573871613,
|
| 1356 |
+
"learning_rate": 2.9166944444444447e-06,
|
| 1357 |
+
"loss": 0.0826,
|
| 1358 |
+
"step": 95000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 2.3875,
|
| 1362 |
+
"grad_norm": 8.53817709867144e-07,
|
| 1363 |
+
"learning_rate": 2.9028055555555557e-06,
|
| 1364 |
+
"loss": 0.0924,
|
| 1365 |
+
"step": 95500
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 2.4,
|
| 1369 |
+
"grad_norm": 0.017462341114878654,
|
| 1370 |
+
"learning_rate": 2.8889166666666667e-06,
|
| 1371 |
+
"loss": 0.0861,
|
| 1372 |
+
"step": 96000
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.4125,
|
| 1376 |
+
"grad_norm": 0.010675615631043911,
|
| 1377 |
+
"learning_rate": 2.875027777777778e-06,
|
| 1378 |
+
"loss": 0.1213,
|
| 1379 |
+
"step": 96500
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 2.425,
|
| 1383 |
+
"grad_norm": 0.0014171106740832329,
|
| 1384 |
+
"learning_rate": 2.861138888888889e-06,
|
| 1385 |
+
"loss": 0.0655,
|
| 1386 |
+
"step": 97000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 2.4375,
|
| 1390 |
+
"grad_norm": 1.936705729121968e-07,
|
| 1391 |
+
"learning_rate": 2.84725e-06,
|
| 1392 |
+
"loss": 0.1056,
|
| 1393 |
+
"step": 97500
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 2.45,
|
| 1397 |
+
"grad_norm": 3.3484088817203883e-06,
|
| 1398 |
+
"learning_rate": 2.8333611111111116e-06,
|
| 1399 |
+
"loss": 0.0603,
|
| 1400 |
+
"step": 98000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 2.4625,
|
| 1404 |
+
"grad_norm": 0.025326106697320938,
|
| 1405 |
+
"learning_rate": 2.8194722222222226e-06,
|
| 1406 |
+
"loss": 0.0809,
|
| 1407 |
+
"step": 98500
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 2.475,
|
| 1411 |
+
"grad_norm": 0.00743254367262125,
|
| 1412 |
+
"learning_rate": 2.8055833333333336e-06,
|
| 1413 |
+
"loss": 0.1357,
|
| 1414 |
+
"step": 99000
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.4875,
|
| 1418 |
+
"grad_norm": 0.01988009549677372,
|
| 1419 |
+
"learning_rate": 2.791694444444445e-06,
|
| 1420 |
+
"loss": 0.0974,
|
| 1421 |
+
"step": 99500
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 2.5,
|
| 1425 |
+
"grad_norm": 0.0004283917951397598,
|
| 1426 |
+
"learning_rate": 2.7778055555555556e-06,
|
| 1427 |
+
"loss": 0.1095,
|
| 1428 |
+
"step": 100000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 2.5125,
|
| 1432 |
+
"grad_norm": 0.005049830302596092,
|
| 1433 |
+
"learning_rate": 2.7639166666666666e-06,
|
| 1434 |
+
"loss": 0.1019,
|
| 1435 |
+
"step": 100500
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 2.525,
|
| 1439 |
+
"grad_norm": 2.7473144655232318e-05,
|
| 1440 |
+
"learning_rate": 2.750027777777778e-06,
|
| 1441 |
+
"loss": 0.0916,
|
| 1442 |
+
"step": 101000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 2.5375,
|
| 1446 |
+
"grad_norm": 0.19764241576194763,
|
| 1447 |
+
"learning_rate": 2.736138888888889e-06,
|
| 1448 |
+
"loss": 0.1209,
|
| 1449 |
+
"step": 101500
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 2.55,
|
| 1453 |
+
"grad_norm": 0.15940068662166595,
|
| 1454 |
+
"learning_rate": 2.72225e-06,
|
| 1455 |
+
"loss": 0.039,
|
| 1456 |
+
"step": 102000
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.5625,
|
| 1460 |
+
"grad_norm": 0.030904898419976234,
|
| 1461 |
+
"learning_rate": 2.7083611111111115e-06,
|
| 1462 |
+
"loss": 0.0581,
|
| 1463 |
+
"step": 102500
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 2.575,
|
| 1467 |
+
"grad_norm": 0.5066580176353455,
|
| 1468 |
+
"learning_rate": 2.6944722222222225e-06,
|
| 1469 |
+
"loss": 0.0971,
|
| 1470 |
+
"step": 103000
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.5875,
|
| 1474 |
+
"grad_norm": 2.1701941577134676e-08,
|
| 1475 |
+
"learning_rate": 2.6805833333333335e-06,
|
| 1476 |
+
"loss": 0.0984,
|
| 1477 |
+
"step": 103500
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 2.6,
|
| 1481 |
+
"grad_norm": 0.0060079507529735565,
|
| 1482 |
+
"learning_rate": 2.666694444444445e-06,
|
| 1483 |
+
"loss": 0.1176,
|
| 1484 |
+
"step": 104000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 2.6125,
|
| 1488 |
+
"grad_norm": 0.01262228935956955,
|
| 1489 |
+
"learning_rate": 2.652805555555556e-06,
|
| 1490 |
+
"loss": 0.0999,
|
| 1491 |
+
"step": 104500
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 2.625,
|
| 1495 |
+
"grad_norm": 0.004787284880876541,
|
| 1496 |
+
"learning_rate": 2.638916666666667e-06,
|
| 1497 |
+
"loss": 0.1031,
|
| 1498 |
+
"step": 105000
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 2.6375,
|
| 1502 |
+
"grad_norm": 0.004941838793456554,
|
| 1503 |
+
"learning_rate": 2.6250277777777776e-06,
|
| 1504 |
+
"loss": 0.1018,
|
| 1505 |
+
"step": 105500
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 2.65,
|
| 1509 |
+
"grad_norm": 2.6329735192121007e-06,
|
| 1510 |
+
"learning_rate": 2.611138888888889e-06,
|
| 1511 |
+
"loss": 0.1081,
|
| 1512 |
+
"step": 106000
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.6625,
|
| 1516 |
+
"grad_norm": 0.028844570741057396,
|
| 1517 |
+
"learning_rate": 2.59725e-06,
|
| 1518 |
+
"loss": 0.0558,
|
| 1519 |
+
"step": 106500
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 2.675,
|
| 1523 |
+
"grad_norm": 1.3074130666979045e-09,
|
| 1524 |
+
"learning_rate": 2.583361111111111e-06,
|
| 1525 |
+
"loss": 0.098,
|
| 1526 |
+
"step": 107000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 2.6875,
|
| 1530 |
+
"grad_norm": 0.1365622878074646,
|
| 1531 |
+
"learning_rate": 2.5694722222222224e-06,
|
| 1532 |
+
"loss": 0.0967,
|
| 1533 |
+
"step": 107500
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 2.7,
|
| 1537 |
+
"grad_norm": 1.7111767647293163e-07,
|
| 1538 |
+
"learning_rate": 2.5555833333333334e-06,
|
| 1539 |
+
"loss": 0.0829,
|
| 1540 |
+
"step": 108000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 2.7125,
|
| 1544 |
+
"grad_norm": 1.4454585652856622e-05,
|
| 1545 |
+
"learning_rate": 2.541694444444445e-06,
|
| 1546 |
+
"loss": 0.1038,
|
| 1547 |
+
"step": 108500
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 2.725,
|
| 1551 |
+
"grad_norm": 0.3521536588668823,
|
| 1552 |
+
"learning_rate": 2.527805555555556e-06,
|
| 1553 |
+
"loss": 0.0797,
|
| 1554 |
+
"step": 109000
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 2.7375,
|
| 1558 |
+
"grad_norm": 0.00021774623019155115,
|
| 1559 |
+
"learning_rate": 2.513916666666667e-06,
|
| 1560 |
+
"loss": 0.1158,
|
| 1561 |
+
"step": 109500
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 2.75,
|
| 1565 |
+
"grad_norm": 0.0001049950733431615,
|
| 1566 |
+
"learning_rate": 2.5000277777777783e-06,
|
| 1567 |
+
"loss": 0.0928,
|
| 1568 |
+
"step": 110000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 2.7625,
|
| 1572 |
+
"grad_norm": 3.3473178007170645e-08,
|
| 1573 |
+
"learning_rate": 2.486138888888889e-06,
|
| 1574 |
+
"loss": 0.0979,
|
| 1575 |
+
"step": 110500
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 2.775,
|
| 1579 |
+
"grad_norm": 0.003986213356256485,
|
| 1580 |
+
"learning_rate": 2.4722500000000003e-06,
|
| 1581 |
+
"loss": 0.0958,
|
| 1582 |
+
"step": 111000
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 2.7875,
|
| 1586 |
+
"grad_norm": 5.052464530308498e-06,
|
| 1587 |
+
"learning_rate": 2.4583611111111114e-06,
|
| 1588 |
+
"loss": 0.0445,
|
| 1589 |
+
"step": 111500
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 2.8,
|
| 1593 |
+
"grad_norm": 1.3031937839969032e-07,
|
| 1594 |
+
"learning_rate": 2.4444722222222224e-06,
|
| 1595 |
+
"loss": 0.0673,
|
| 1596 |
+
"step": 112000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 2.8125,
|
| 1600 |
+
"grad_norm": 0.013562874868512154,
|
| 1601 |
+
"learning_rate": 2.4305833333333334e-06,
|
| 1602 |
+
"loss": 0.0778,
|
| 1603 |
+
"step": 112500
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 2.825,
|
| 1607 |
+
"grad_norm": 1.478714057157049e-05,
|
| 1608 |
+
"learning_rate": 2.4166944444444444e-06,
|
| 1609 |
+
"loss": 0.0978,
|
| 1610 |
+
"step": 113000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 2.8375,
|
| 1614 |
+
"grad_norm": 0.15072928369045258,
|
| 1615 |
+
"learning_rate": 2.402805555555556e-06,
|
| 1616 |
+
"loss": 0.036,
|
| 1617 |
+
"step": 113500
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 2.85,
|
| 1621 |
+
"grad_norm": 0.15145017206668854,
|
| 1622 |
+
"learning_rate": 2.388916666666667e-06,
|
| 1623 |
+
"loss": 0.0779,
|
| 1624 |
+
"step": 114000
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 2.8625,
|
| 1628 |
+
"grad_norm": 0.0001674382365308702,
|
| 1629 |
+
"learning_rate": 2.3750277777777783e-06,
|
| 1630 |
+
"loss": 0.1006,
|
| 1631 |
+
"step": 114500
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 2.875,
|
| 1635 |
+
"grad_norm": 0.0002086051827063784,
|
| 1636 |
+
"learning_rate": 2.3611388888888893e-06,
|
| 1637 |
+
"loss": 0.0734,
|
| 1638 |
+
"step": 115000
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 2.8875,
|
| 1642 |
+
"grad_norm": 164.32354736328125,
|
| 1643 |
+
"learning_rate": 2.3472500000000003e-06,
|
| 1644 |
+
"loss": 0.0729,
|
| 1645 |
+
"step": 115500
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 2.9,
|
| 1649 |
+
"grad_norm": 0.012316230684518814,
|
| 1650 |
+
"learning_rate": 2.3333611111111113e-06,
|
| 1651 |
+
"loss": 0.0984,
|
| 1652 |
+
"step": 116000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 2.9125,
|
| 1656 |
+
"grad_norm": 0.32554274797439575,
|
| 1657 |
+
"learning_rate": 2.3194722222222223e-06,
|
| 1658 |
+
"loss": 0.0906,
|
| 1659 |
+
"step": 116500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 2.925,
|
| 1663 |
+
"grad_norm": 7.992193218342436e-07,
|
| 1664 |
+
"learning_rate": 2.3055833333333337e-06,
|
| 1665 |
+
"loss": 0.0891,
|
| 1666 |
+
"step": 117000
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 2.9375,
|
| 1670 |
+
"grad_norm": 0.001029053470119834,
|
| 1671 |
+
"learning_rate": 2.2916944444444447e-06,
|
| 1672 |
+
"loss": 0.104,
|
| 1673 |
+
"step": 117500
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 2.95,
|
| 1677 |
+
"grad_norm": 0.0013586468994617462,
|
| 1678 |
+
"learning_rate": 2.2778055555555557e-06,
|
| 1679 |
+
"loss": 0.04,
|
| 1680 |
+
"step": 118000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 2.9625,
|
| 1684 |
+
"grad_norm": 5.0012647989206016e-05,
|
| 1685 |
+
"learning_rate": 2.2639166666666667e-06,
|
| 1686 |
+
"loss": 0.0689,
|
| 1687 |
+
"step": 118500
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 2.975,
|
| 1691 |
+
"grad_norm": 0.045165080577135086,
|
| 1692 |
+
"learning_rate": 2.2500277777777778e-06,
|
| 1693 |
+
"loss": 0.0758,
|
| 1694 |
+
"step": 119000
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 2.9875,
|
| 1698 |
+
"grad_norm": 4.196267582301516e-07,
|
| 1699 |
+
"learning_rate": 2.236138888888889e-06,
|
| 1700 |
+
"loss": 0.0596,
|
| 1701 |
+
"step": 119500
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 3.0,
|
| 1705 |
+
"grad_norm": 0.00039908240432851017,
|
| 1706 |
+
"learning_rate": 2.22225e-06,
|
| 1707 |
+
"loss": 0.0775,
|
| 1708 |
+
"step": 120000
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 3.0,
|
| 1712 |
+
"eval_accuracy": 0.9856,
|
| 1713 |
+
"eval_loss": 0.09523609280586243,
|
| 1714 |
+
"eval_runtime": 70.8442,
|
| 1715 |
+
"eval_samples_per_second": 282.31,
|
| 1716 |
+
"eval_steps_per_second": 70.577,
|
| 1717 |
+
"step": 120000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.0125,
|
| 1721 |
+
"grad_norm": 0.0008493961067870259,
|
| 1722 |
+
"learning_rate": 2.208361111111111e-06,
|
| 1723 |
+
"loss": 0.0756,
|
| 1724 |
+
"step": 120500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.025,
|
| 1728 |
+
"grad_norm": 3.1318024412030354e-05,
|
| 1729 |
+
"learning_rate": 2.1944722222222222e-06,
|
| 1730 |
+
"loss": 0.061,
|
| 1731 |
+
"step": 121000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.0375,
|
| 1735 |
+
"grad_norm": 7.05946590073836e-08,
|
| 1736 |
+
"learning_rate": 2.1805833333333336e-06,
|
| 1737 |
+
"loss": 0.032,
|
| 1738 |
+
"step": 121500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.05,
|
| 1742 |
+
"grad_norm": 2.9236124987619405e-07,
|
| 1743 |
+
"learning_rate": 2.1666944444444447e-06,
|
| 1744 |
+
"loss": 0.049,
|
| 1745 |
+
"step": 122000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.0625,
|
| 1749 |
+
"grad_norm": 3.1436852623301093e-06,
|
| 1750 |
+
"learning_rate": 2.1528055555555557e-06,
|
| 1751 |
+
"loss": 0.0465,
|
| 1752 |
+
"step": 122500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.075,
|
| 1756 |
+
"grad_norm": 7.781285660257709e-08,
|
| 1757 |
+
"learning_rate": 2.138916666666667e-06,
|
| 1758 |
+
"loss": 0.0702,
|
| 1759 |
+
"step": 123000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.0875,
|
| 1763 |
+
"grad_norm": 2.085931782858097e-07,
|
| 1764 |
+
"learning_rate": 2.125027777777778e-06,
|
| 1765 |
+
"loss": 0.0448,
|
| 1766 |
+
"step": 123500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 3.1,
|
| 1770 |
+
"grad_norm": 1024.1812744140625,
|
| 1771 |
+
"learning_rate": 2.111138888888889e-06,
|
| 1772 |
+
"loss": 0.0898,
|
| 1773 |
+
"step": 124000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 3.1125,
|
| 1777 |
+
"grad_norm": 1.351660188220194e-07,
|
| 1778 |
+
"learning_rate": 2.09725e-06,
|
| 1779 |
+
"loss": 0.0431,
|
| 1780 |
+
"step": 124500
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 3.125,
|
| 1784 |
+
"grad_norm": 0.001149096991866827,
|
| 1785 |
+
"learning_rate": 2.083361111111111e-06,
|
| 1786 |
+
"loss": 0.0486,
|
| 1787 |
+
"step": 125000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 3.1375,
|
| 1791 |
+
"grad_norm": 2.472369851602707e-05,
|
| 1792 |
+
"learning_rate": 2.0694722222222226e-06,
|
| 1793 |
+
"loss": 0.0539,
|
| 1794 |
+
"step": 125500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 3.15,
|
| 1798 |
+
"grad_norm": 6.923924615875876e-07,
|
| 1799 |
+
"learning_rate": 2.0555833333333336e-06,
|
| 1800 |
+
"loss": 0.0751,
|
| 1801 |
+
"step": 126000
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 3.1625,
|
| 1805 |
+
"grad_norm": 3.3076943850574025e-08,
|
| 1806 |
+
"learning_rate": 2.0416944444444446e-06,
|
| 1807 |
+
"loss": 0.0216,
|
| 1808 |
+
"step": 126500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 3.175,
|
| 1812 |
+
"grad_norm": 1.5224570404370752e-07,
|
| 1813 |
+
"learning_rate": 2.0278055555555556e-06,
|
| 1814 |
+
"loss": 0.0425,
|
| 1815 |
+
"step": 127000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 3.1875,
|
| 1819 |
+
"grad_norm": 2.331158430024516e-05,
|
| 1820 |
+
"learning_rate": 2.0139166666666666e-06,
|
| 1821 |
+
"loss": 0.0568,
|
| 1822 |
+
"step": 127500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 3.2,
|
| 1826 |
+
"grad_norm": 0.0012923574540764093,
|
| 1827 |
+
"learning_rate": 2.000027777777778e-06,
|
| 1828 |
+
"loss": 0.0467,
|
| 1829 |
+
"step": 128000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 3.2125,
|
| 1833 |
+
"grad_norm": 0.0013359450967982411,
|
| 1834 |
+
"learning_rate": 1.986138888888889e-06,
|
| 1835 |
+
"loss": 0.0781,
|
| 1836 |
+
"step": 128500
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 3.225,
|
| 1840 |
+
"grad_norm": 0.007183945272117853,
|
| 1841 |
+
"learning_rate": 1.9722500000000005e-06,
|
| 1842 |
+
"loss": 0.0328,
|
| 1843 |
+
"step": 129000
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 3.2375,
|
| 1847 |
+
"grad_norm": 2.5895735234371386e-05,
|
| 1848 |
+
"learning_rate": 1.958361111111111e-06,
|
| 1849 |
+
"loss": 0.0565,
|
| 1850 |
+
"step": 129500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 3.25,
|
| 1854 |
+
"grad_norm": 0.003281608922407031,
|
| 1855 |
+
"learning_rate": 1.9444722222222225e-06,
|
| 1856 |
+
"loss": 0.0573,
|
| 1857 |
+
"step": 130000
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 3.2625,
|
| 1861 |
+
"grad_norm": 7.837111581920908e-08,
|
| 1862 |
+
"learning_rate": 1.9305833333333335e-06,
|
| 1863 |
+
"loss": 0.0129,
|
| 1864 |
+
"step": 130500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 3.275,
|
| 1868 |
+
"grad_norm": 1.010946107271593e-06,
|
| 1869 |
+
"learning_rate": 1.9166944444444445e-06,
|
| 1870 |
+
"loss": 0.0716,
|
| 1871 |
+
"step": 131000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 3.2875,
|
| 1875 |
+
"grad_norm": 1.7364520317642018e-05,
|
| 1876 |
+
"learning_rate": 1.9028055555555557e-06,
|
| 1877 |
+
"loss": 0.0342,
|
| 1878 |
+
"step": 131500
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 3.3,
|
| 1882 |
+
"grad_norm": 2.711913194008986e-10,
|
| 1883 |
+
"learning_rate": 1.888916666666667e-06,
|
| 1884 |
+
"loss": 0.0652,
|
| 1885 |
+
"step": 132000
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 3.3125,
|
| 1889 |
+
"grad_norm": 2.3724902348476462e-05,
|
| 1890 |
+
"learning_rate": 1.8750277777777777e-06,
|
| 1891 |
+
"loss": 0.0512,
|
| 1892 |
+
"step": 132500
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 3.325,
|
| 1896 |
+
"grad_norm": 1.995879850280602e-10,
|
| 1897 |
+
"learning_rate": 1.861138888888889e-06,
|
| 1898 |
+
"loss": 0.051,
|
| 1899 |
+
"step": 133000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 3.3375,
|
| 1903 |
+
"grad_norm": 1.3715705238226406e-09,
|
| 1904 |
+
"learning_rate": 1.8472500000000002e-06,
|
| 1905 |
+
"loss": 0.0173,
|
| 1906 |
+
"step": 133500
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 3.35,
|
| 1910 |
+
"grad_norm": 2.5874431486272442e-08,
|
| 1911 |
+
"learning_rate": 1.8333611111111114e-06,
|
| 1912 |
+
"loss": 0.0265,
|
| 1913 |
+
"step": 134000
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 3.3625,
|
| 1917 |
+
"grad_norm": 9.791951924853493e-07,
|
| 1918 |
+
"learning_rate": 1.8194722222222224e-06,
|
| 1919 |
+
"loss": 0.0639,
|
| 1920 |
+
"step": 134500
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 3.375,
|
| 1924 |
+
"grad_norm": 1.8817563826600825e-11,
|
| 1925 |
+
"learning_rate": 1.8055833333333334e-06,
|
| 1926 |
+
"loss": 0.0213,
|
| 1927 |
+
"step": 135000
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 3.3875,
|
| 1931 |
+
"grad_norm": 1.4598784900954342e-06,
|
| 1932 |
+
"learning_rate": 1.7916944444444444e-06,
|
| 1933 |
+
"loss": 0.0356,
|
| 1934 |
+
"step": 135500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 3.4,
|
| 1938 |
+
"grad_norm": 4.592724280882976e-07,
|
| 1939 |
+
"learning_rate": 1.7778055555555557e-06,
|
| 1940 |
+
"loss": 0.0767,
|
| 1941 |
+
"step": 136000
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 3.4125,
|
| 1945 |
+
"grad_norm": 1.918931502586929e-07,
|
| 1946 |
+
"learning_rate": 1.7639166666666669e-06,
|
| 1947 |
+
"loss": 0.0157,
|
| 1948 |
+
"step": 136500
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 3.425,
|
| 1952 |
+
"grad_norm": 5.836361651745392e-07,
|
| 1953 |
+
"learning_rate": 1.750027777777778e-06,
|
| 1954 |
+
"loss": 0.0424,
|
| 1955 |
+
"step": 137000
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 3.4375,
|
| 1959 |
+
"grad_norm": 5.298739324643975e-06,
|
| 1960 |
+
"learning_rate": 1.736138888888889e-06,
|
| 1961 |
+
"loss": 0.0655,
|
| 1962 |
+
"step": 137500
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 3.45,
|
| 1966 |
+
"grad_norm": 7.815018761903048e-05,
|
| 1967 |
+
"learning_rate": 1.7222500000000001e-06,
|
| 1968 |
+
"loss": 0.0194,
|
| 1969 |
+
"step": 138000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.4625,
|
| 1973 |
+
"grad_norm": 0.0004915116005577147,
|
| 1974 |
+
"learning_rate": 1.7083611111111111e-06,
|
| 1975 |
+
"loss": 0.0949,
|
| 1976 |
+
"step": 138500
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.475,
|
| 1980 |
+
"grad_norm": 0.002174359979107976,
|
| 1981 |
+
"learning_rate": 1.6944722222222223e-06,
|
| 1982 |
+
"loss": 0.0587,
|
| 1983 |
+
"step": 139000
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.4875,
|
| 1987 |
+
"grad_norm": 0.0017873262986540794,
|
| 1988 |
+
"learning_rate": 1.6805833333333336e-06,
|
| 1989 |
+
"loss": 0.0659,
|
| 1990 |
+
"step": 139500
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.5,
|
| 1994 |
+
"grad_norm": 0.00011011360766133294,
|
| 1995 |
+
"learning_rate": 1.6666944444444446e-06,
|
| 1996 |
+
"loss": 0.0454,
|
| 1997 |
+
"step": 140000
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.5125,
|
| 2001 |
+
"grad_norm": 5.8530247770249844e-05,
|
| 2002 |
+
"learning_rate": 1.6528055555555556e-06,
|
| 2003 |
+
"loss": 0.0209,
|
| 2004 |
+
"step": 140500
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.525,
|
| 2008 |
+
"grad_norm": 0.0005035810172557831,
|
| 2009 |
+
"learning_rate": 1.6389166666666668e-06,
|
| 2010 |
+
"loss": 0.0575,
|
| 2011 |
+
"step": 141000
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.5375,
|
| 2015 |
+
"grad_norm": 7.48745515011251e-05,
|
| 2016 |
+
"learning_rate": 1.6250277777777778e-06,
|
| 2017 |
+
"loss": 0.0354,
|
| 2018 |
+
"step": 141500
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.55,
|
| 2022 |
+
"grad_norm": 2.5532403014949523e-05,
|
| 2023 |
+
"learning_rate": 1.611138888888889e-06,
|
| 2024 |
+
"loss": 0.0304,
|
| 2025 |
+
"step": 142000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.5625,
|
| 2029 |
+
"grad_norm": 8.711179543752223e-05,
|
| 2030 |
+
"learning_rate": 1.5972500000000003e-06,
|
| 2031 |
+
"loss": 0.0483,
|
| 2032 |
+
"step": 142500
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.575,
|
| 2036 |
+
"grad_norm": 1.006107552825597e-08,
|
| 2037 |
+
"learning_rate": 1.5833611111111113e-06,
|
| 2038 |
+
"loss": 0.04,
|
| 2039 |
+
"step": 143000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.5875,
|
| 2043 |
+
"grad_norm": 0.00023778372269589454,
|
| 2044 |
+
"learning_rate": 1.5694722222222223e-06,
|
| 2045 |
+
"loss": 0.0633,
|
| 2046 |
+
"step": 143500
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.6,
|
| 2050 |
+
"grad_norm": 0.09549811482429504,
|
| 2051 |
+
"learning_rate": 1.5555833333333335e-06,
|
| 2052 |
+
"loss": 0.0593,
|
| 2053 |
+
"step": 144000
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.6125,
|
| 2057 |
+
"grad_norm": 0.0007943428936414421,
|
| 2058 |
+
"learning_rate": 1.5416944444444445e-06,
|
| 2059 |
+
"loss": 0.0664,
|
| 2060 |
+
"step": 144500
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.625,
|
| 2064 |
+
"grad_norm": 187.2425994873047,
|
| 2065 |
+
"learning_rate": 1.5278055555555557e-06,
|
| 2066 |
+
"loss": 0.0574,
|
| 2067 |
+
"step": 145000
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.6375,
|
| 2071 |
+
"grad_norm": 8.670787792652845e-06,
|
| 2072 |
+
"learning_rate": 1.513916666666667e-06,
|
| 2073 |
+
"loss": 0.0284,
|
| 2074 |
+
"step": 145500
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.65,
|
| 2078 |
+
"grad_norm": 6.026363553246483e-05,
|
| 2079 |
+
"learning_rate": 1.500027777777778e-06,
|
| 2080 |
+
"loss": 0.0188,
|
| 2081 |
+
"step": 146000
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.6625,
|
| 2085 |
+
"grad_norm": 2.375714514002425e-09,
|
| 2086 |
+
"learning_rate": 1.486138888888889e-06,
|
| 2087 |
+
"loss": 0.0257,
|
| 2088 |
+
"step": 146500
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.675,
|
| 2092 |
+
"grad_norm": 6.045565532986075e-05,
|
| 2093 |
+
"learning_rate": 1.47225e-06,
|
| 2094 |
+
"loss": 0.065,
|
| 2095 |
+
"step": 147000
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.6875,
|
| 2099 |
+
"grad_norm": 0.0009056538692675531,
|
| 2100 |
+
"learning_rate": 1.4583611111111112e-06,
|
| 2101 |
+
"loss": 0.0666,
|
| 2102 |
+
"step": 147500
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.7,
|
| 2106 |
+
"grad_norm": 0.013074205256998539,
|
| 2107 |
+
"learning_rate": 1.4444722222222224e-06,
|
| 2108 |
+
"loss": 0.0439,
|
| 2109 |
+
"step": 148000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.7125,
|
| 2113 |
+
"grad_norm": 0.0008992800139822066,
|
| 2114 |
+
"learning_rate": 1.4305833333333336e-06,
|
| 2115 |
+
"loss": 0.0542,
|
| 2116 |
+
"step": 148500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.725,
|
| 2120 |
+
"grad_norm": 1.5302573430631128e-08,
|
| 2121 |
+
"learning_rate": 1.4166944444444444e-06,
|
| 2122 |
+
"loss": 0.059,
|
| 2123 |
+
"step": 149000
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.7375,
|
| 2127 |
+
"grad_norm": 0.015462064184248447,
|
| 2128 |
+
"learning_rate": 1.4028055555555556e-06,
|
| 2129 |
+
"loss": 0.0444,
|
| 2130 |
+
"step": 149500
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.75,
|
| 2134 |
+
"grad_norm": 2.3089251044439152e-05,
|
| 2135 |
+
"learning_rate": 1.3889166666666667e-06,
|
| 2136 |
+
"loss": 0.0488,
|
| 2137 |
+
"step": 150000
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.7625,
|
| 2141 |
+
"grad_norm": 0.029342608526349068,
|
| 2142 |
+
"learning_rate": 1.3750277777777779e-06,
|
| 2143 |
+
"loss": 0.0387,
|
| 2144 |
+
"step": 150500
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.775,
|
| 2148 |
+
"grad_norm": 0.16643966734409332,
|
| 2149 |
+
"learning_rate": 1.361138888888889e-06,
|
| 2150 |
+
"loss": 0.0496,
|
| 2151 |
+
"step": 151000
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.7875,
|
| 2155 |
+
"grad_norm": 3.40620975869399e-10,
|
| 2156 |
+
"learning_rate": 1.3472500000000003e-06,
|
| 2157 |
+
"loss": 0.037,
|
| 2158 |
+
"step": 151500
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.8,
|
| 2162 |
+
"grad_norm": 1.0341510980893531e-09,
|
| 2163 |
+
"learning_rate": 1.3333611111111111e-06,
|
| 2164 |
+
"loss": 0.0434,
|
| 2165 |
+
"step": 152000
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.8125,
|
| 2169 |
+
"grad_norm": 6.396485696313903e-05,
|
| 2170 |
+
"learning_rate": 1.3194722222222223e-06,
|
| 2171 |
+
"loss": 0.0924,
|
| 2172 |
+
"step": 152500
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.825,
|
| 2176 |
+
"grad_norm": 2.9705923831890857e-10,
|
| 2177 |
+
"learning_rate": 1.3055833333333333e-06,
|
| 2178 |
+
"loss": 0.0319,
|
| 2179 |
+
"step": 153000
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.8375,
|
| 2183 |
+
"grad_norm": 0.00013349021901376545,
|
| 2184 |
+
"learning_rate": 1.2916944444444446e-06,
|
| 2185 |
+
"loss": 0.028,
|
| 2186 |
+
"step": 153500
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.85,
|
| 2190 |
+
"grad_norm": 1.4059290620593856e-08,
|
| 2191 |
+
"learning_rate": 1.2778055555555558e-06,
|
| 2192 |
+
"loss": 0.0219,
|
| 2193 |
+
"step": 154000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 3.8625,
|
| 2197 |
+
"grad_norm": 1.5280780871762545e-06,
|
| 2198 |
+
"learning_rate": 1.263916666666667e-06,
|
| 2199 |
+
"loss": 0.0522,
|
| 2200 |
+
"step": 154500
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 3.875,
|
| 2204 |
+
"grad_norm": 4.033211264964365e-10,
|
| 2205 |
+
"learning_rate": 1.2500277777777778e-06,
|
| 2206 |
+
"loss": 0.0331,
|
| 2207 |
+
"step": 155000
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 3.8875,
|
| 2211 |
+
"grad_norm": 0.004559781868010759,
|
| 2212 |
+
"learning_rate": 1.236138888888889e-06,
|
| 2213 |
+
"loss": 0.0628,
|
| 2214 |
+
"step": 155500
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 3.9,
|
| 2218 |
+
"grad_norm": 4.622693813871592e-05,
|
| 2219 |
+
"learning_rate": 1.22225e-06,
|
| 2220 |
+
"loss": 0.0114,
|
| 2221 |
+
"step": 156000
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 3.9125,
|
| 2225 |
+
"grad_norm": 3.810692703609675e-07,
|
| 2226 |
+
"learning_rate": 1.2083611111111112e-06,
|
| 2227 |
+
"loss": 0.0466,
|
| 2228 |
+
"step": 156500
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 3.925,
|
| 2232 |
+
"grad_norm": 9.439718269277364e-07,
|
| 2233 |
+
"learning_rate": 1.1944722222222223e-06,
|
| 2234 |
+
"loss": 0.0058,
|
| 2235 |
+
"step": 157000
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 3.9375,
|
| 2239 |
+
"grad_norm": 2.499190010496477e-08,
|
| 2240 |
+
"learning_rate": 1.1805833333333335e-06,
|
| 2241 |
+
"loss": 0.0573,
|
| 2242 |
+
"step": 157500
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 3.95,
|
| 2246 |
+
"grad_norm": 2.330422574914337e-07,
|
| 2247 |
+
"learning_rate": 1.1666944444444447e-06,
|
| 2248 |
+
"loss": 0.0372,
|
| 2249 |
+
"step": 158000
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 3.9625,
|
| 2253 |
+
"grad_norm": 2.389850806139293e-08,
|
| 2254 |
+
"learning_rate": 1.1528055555555557e-06,
|
| 2255 |
+
"loss": 0.0124,
|
| 2256 |
+
"step": 158500
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.975,
|
| 2260 |
+
"grad_norm": 1.5795259544759688e-09,
|
| 2261 |
+
"learning_rate": 1.1389166666666667e-06,
|
| 2262 |
+
"loss": 0.0521,
|
| 2263 |
+
"step": 159000
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 3.9875,
|
| 2267 |
+
"grad_norm": 1.8120088762785258e-09,
|
| 2268 |
+
"learning_rate": 1.1250277777777777e-06,
|
| 2269 |
+
"loss": 0.0346,
|
| 2270 |
+
"step": 159500
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 4.0,
|
| 2274 |
+
"grad_norm": 7.511826161099577e-10,
|
| 2275 |
+
"learning_rate": 1.111138888888889e-06,
|
| 2276 |
+
"loss": 0.0671,
|
| 2277 |
+
"step": 160000
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 4.0,
|
| 2281 |
+
"eval_accuracy": 0.9812,
|
| 2282 |
+
"eval_loss": 0.14172975718975067,
|
| 2283 |
+
"eval_runtime": 71.1854,
|
| 2284 |
+
"eval_samples_per_second": 280.957,
|
| 2285 |
+
"eval_steps_per_second": 70.239,
|
| 2286 |
+
"step": 160000
|
| 2287 |
+
}
|
| 2288 |
+
],
|
| 2289 |
+
"logging_steps": 500,
|
| 2290 |
+
"max_steps": 200000,
|
| 2291 |
+
"num_input_tokens_seen": 0,
|
| 2292 |
+
"num_train_epochs": 5,
|
| 2293 |
+
"save_steps": 500,
|
| 2294 |
+
"stateful_callbacks": {
|
| 2295 |
+
"TrainerControl": {
|
| 2296 |
+
"args": {
|
| 2297 |
+
"should_epoch_stop": false,
|
| 2298 |
+
"should_evaluate": false,
|
| 2299 |
+
"should_log": false,
|
| 2300 |
+
"should_save": true,
|
| 2301 |
+
"should_training_stop": false
|
| 2302 |
+
},
|
| 2303 |
+
"attributes": {}
|
| 2304 |
+
}
|
| 2305 |
+
},
|
| 2306 |
+
"total_flos": 2.680175866281984e+19,
|
| 2307 |
+
"train_batch_size": 2,
|
| 2308 |
+
"trial_name": null,
|
| 2309 |
+
"trial_params": null
|
| 2310 |
+
}
|
checkpoint-160000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741392110574dc74949de4995bd5bf124fe829c8e6a7e3937ee5dcd15b77b966
|
| 3 |
+
size 5265
|
checkpoint-200000/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"dtype": "float32",
|
| 6 |
+
"id2label": {
|
| 7 |
+
"0": "ai",
|
| 8 |
+
"1": "hum"
|
| 9 |
+
},
|
| 10 |
+
"initializer_factor": 1.0,
|
| 11 |
+
"label2id": {
|
| 12 |
+
"ai": 0,
|
| 13 |
+
"hum": 1
|
| 14 |
+
},
|
| 15 |
+
"model_type": "siglip",
|
| 16 |
+
"problem_type": "single_label_classification",
|
| 17 |
+
"text_config": {
|
| 18 |
+
"attention_dropout": 0.0,
|
| 19 |
+
"dtype": "float32",
|
| 20 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"layer_norm_eps": 1e-06,
|
| 24 |
+
"max_position_embeddings": 64,
|
| 25 |
+
"model_type": "siglip_text_model",
|
| 26 |
+
"num_attention_heads": 12,
|
| 27 |
+
"num_hidden_layers": 12,
|
| 28 |
+
"projection_size": 768,
|
| 29 |
+
"vocab_size": 256000
|
| 30 |
+
},
|
| 31 |
+
"transformers_version": "5.0.0.dev0",
|
| 32 |
+
"use_cache": false,
|
| 33 |
+
"vision_config": {
|
| 34 |
+
"attention_dropout": 0.0,
|
| 35 |
+
"dtype": "float32",
|
| 36 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 37 |
+
"hidden_size": 768,
|
| 38 |
+
"image_size": 224,
|
| 39 |
+
"intermediate_size": 3072,
|
| 40 |
+
"layer_norm_eps": 1e-06,
|
| 41 |
+
"model_type": "siglip_vision_model",
|
| 42 |
+
"num_attention_heads": 12,
|
| 43 |
+
"num_channels": 3,
|
| 44 |
+
"num_hidden_layers": 12,
|
| 45 |
+
"patch_size": 16
|
| 46 |
+
}
|
| 47 |
+
}
|
checkpoint-200000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:510e152b05df42e0fbd31fea37a63bf00b6be2749dec448b0c95810c67a7c94f
|
| 3 |
+
size 371567992
|
checkpoint-200000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a9aa17fdd02c36aeecd0da602f5c0a8e941d11024c102a5e65e02f2acace88d
|
| 3 |
+
size 686558987
|
checkpoint-200000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3cc5713982198cf768e63fa5a0aafdc20acc22881d3590f13ed0f6f7971eb24f
|
| 3 |
+
size 14645
|
checkpoint-200000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90fa5d1774c8f30ebc14c673ee34fbfbc3fe330442cc80f54ddd38e30b6f792f
|
| 3 |
+
size 1465
|
checkpoint-200000/trainer_state.json
ADDED
|
@@ -0,0 +1,2879 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 200000,
|
| 3 |
+
"best_metric": 0.98965,
|
| 4 |
+
"best_model_checkpoint": "ai-vs-hum/checkpoint-200000",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 200000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0125,
|
| 14 |
+
"grad_norm": 57.61789321899414,
|
| 15 |
+
"learning_rate": 1.2475e-07,
|
| 16 |
+
"loss": 1.0591,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.025,
|
| 21 |
+
"grad_norm": 69.49858093261719,
|
| 22 |
+
"learning_rate": 2.4975e-07,
|
| 23 |
+
"loss": 0.5393,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0375,
|
| 28 |
+
"grad_norm": 1023.806640625,
|
| 29 |
+
"learning_rate": 3.7475000000000004e-07,
|
| 30 |
+
"loss": 0.4548,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05,
|
| 35 |
+
"grad_norm": 777.0330810546875,
|
| 36 |
+
"learning_rate": 4.997500000000001e-07,
|
| 37 |
+
"loss": 0.6269,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0625,
|
| 42 |
+
"grad_norm": 0.08837065100669861,
|
| 43 |
+
"learning_rate": 6.2475e-07,
|
| 44 |
+
"loss": 0.5631,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.075,
|
| 49 |
+
"grad_norm": 0.024585120379924774,
|
| 50 |
+
"learning_rate": 7.497500000000001e-07,
|
| 51 |
+
"loss": 0.5938,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0875,
|
| 56 |
+
"grad_norm": 0.012241708114743233,
|
| 57 |
+
"learning_rate": 8.747500000000001e-07,
|
| 58 |
+
"loss": 0.5345,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1,
|
| 63 |
+
"grad_norm": 0.37598320841789246,
|
| 64 |
+
"learning_rate": 9.9975e-07,
|
| 65 |
+
"loss": 0.4773,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1125,
|
| 70 |
+
"grad_norm": 0.11799755692481995,
|
| 71 |
+
"learning_rate": 1.1247500000000001e-06,
|
| 72 |
+
"loss": 0.4258,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.125,
|
| 77 |
+
"grad_norm": 0.034392450004816055,
|
| 78 |
+
"learning_rate": 1.2497500000000002e-06,
|
| 79 |
+
"loss": 0.578,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1375,
|
| 84 |
+
"grad_norm": 775.758056640625,
|
| 85 |
+
"learning_rate": 1.3747500000000003e-06,
|
| 86 |
+
"loss": 0.5597,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.15,
|
| 91 |
+
"grad_norm": 0.01906261406838894,
|
| 92 |
+
"learning_rate": 1.4997500000000001e-06,
|
| 93 |
+
"loss": 0.4832,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1625,
|
| 98 |
+
"grad_norm": 0.016003865748643875,
|
| 99 |
+
"learning_rate": 1.6247500000000002e-06,
|
| 100 |
+
"loss": 0.4621,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.175,
|
| 105 |
+
"grad_norm": 0.052252043038606644,
|
| 106 |
+
"learning_rate": 1.74975e-06,
|
| 107 |
+
"loss": 0.4056,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1875,
|
| 112 |
+
"grad_norm": 0.008017883636057377,
|
| 113 |
+
"learning_rate": 1.8747500000000001e-06,
|
| 114 |
+
"loss": 0.4268,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2,
|
| 119 |
+
"grad_norm": 0.04491295665502548,
|
| 120 |
+
"learning_rate": 1.99975e-06,
|
| 121 |
+
"loss": 0.4946,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2125,
|
| 126 |
+
"grad_norm": 0.24870972335338593,
|
| 127 |
+
"learning_rate": 2.1247500000000003e-06,
|
| 128 |
+
"loss": 0.3448,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.225,
|
| 133 |
+
"grad_norm": 0.39745479822158813,
|
| 134 |
+
"learning_rate": 2.2497500000000004e-06,
|
| 135 |
+
"loss": 0.5246,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2375,
|
| 140 |
+
"grad_norm": 0.1810966432094574,
|
| 141 |
+
"learning_rate": 2.37475e-06,
|
| 142 |
+
"loss": 0.3799,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.25,
|
| 147 |
+
"grad_norm": 0.03164760395884514,
|
| 148 |
+
"learning_rate": 2.49975e-06,
|
| 149 |
+
"loss": 0.3887,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.2625,
|
| 154 |
+
"grad_norm": 22.039918899536133,
|
| 155 |
+
"learning_rate": 2.62475e-06,
|
| 156 |
+
"loss": 0.3956,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.275,
|
| 161 |
+
"grad_norm": 0.06578505039215088,
|
| 162 |
+
"learning_rate": 2.7497500000000007e-06,
|
| 163 |
+
"loss": 0.4013,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2875,
|
| 168 |
+
"grad_norm": 0.016265813261270523,
|
| 169 |
+
"learning_rate": 2.87475e-06,
|
| 170 |
+
"loss": 0.3442,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3,
|
| 175 |
+
"grad_norm": 0.016749408096075058,
|
| 176 |
+
"learning_rate": 2.99975e-06,
|
| 177 |
+
"loss": 0.404,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3125,
|
| 182 |
+
"grad_norm": 0.0033249990083277225,
|
| 183 |
+
"learning_rate": 3.1247500000000004e-06,
|
| 184 |
+
"loss": 0.3088,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.325,
|
| 189 |
+
"grad_norm": 0.3450980484485626,
|
| 190 |
+
"learning_rate": 3.2497500000000005e-06,
|
| 191 |
+
"loss": 0.4177,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3375,
|
| 196 |
+
"grad_norm": 0.01602231152355671,
|
| 197 |
+
"learning_rate": 3.3747500000000006e-06,
|
| 198 |
+
"loss": 0.351,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.35,
|
| 203 |
+
"grad_norm": 0.00211876118555665,
|
| 204 |
+
"learning_rate": 3.4997500000000002e-06,
|
| 205 |
+
"loss": 0.3653,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3625,
|
| 210 |
+
"grad_norm": 0.023136630654335022,
|
| 211 |
+
"learning_rate": 3.6247500000000003e-06,
|
| 212 |
+
"loss": 0.3439,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.375,
|
| 217 |
+
"grad_norm": 0.002042749198153615,
|
| 218 |
+
"learning_rate": 3.7497500000000004e-06,
|
| 219 |
+
"loss": 0.3052,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.3875,
|
| 224 |
+
"grad_norm": 0.023861056193709373,
|
| 225 |
+
"learning_rate": 3.8747500000000005e-06,
|
| 226 |
+
"loss": 0.3397,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4,
|
| 231 |
+
"grad_norm": 0.13855504989624023,
|
| 232 |
+
"learning_rate": 3.999750000000001e-06,
|
| 233 |
+
"loss": 0.2901,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.4125,
|
| 238 |
+
"grad_norm": 0.05347450450062752,
|
| 239 |
+
"learning_rate": 4.12475e-06,
|
| 240 |
+
"loss": 0.3356,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.425,
|
| 245 |
+
"grad_norm": 186.85972595214844,
|
| 246 |
+
"learning_rate": 4.24975e-06,
|
| 247 |
+
"loss": 0.4037,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.4375,
|
| 252 |
+
"grad_norm": 0.009446755982935429,
|
| 253 |
+
"learning_rate": 4.374750000000001e-06,
|
| 254 |
+
"loss": 0.3582,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.45,
|
| 259 |
+
"grad_norm": 0.02892524003982544,
|
| 260 |
+
"learning_rate": 4.49975e-06,
|
| 261 |
+
"loss": 0.2834,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4625,
|
| 266 |
+
"grad_norm": 0.024663692340254784,
|
| 267 |
+
"learning_rate": 4.624750000000001e-06,
|
| 268 |
+
"loss": 0.3806,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.475,
|
| 273 |
+
"grad_norm": 0.17350807785987854,
|
| 274 |
+
"learning_rate": 4.7497500000000006e-06,
|
| 275 |
+
"loss": 0.3418,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.4875,
|
| 280 |
+
"grad_norm": 0.02292953059077263,
|
| 281 |
+
"learning_rate": 4.87475e-06,
|
| 282 |
+
"loss": 0.3369,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.5,
|
| 287 |
+
"grad_norm": 0.014667114242911339,
|
| 288 |
+
"learning_rate": 4.999750000000001e-06,
|
| 289 |
+
"loss": 0.3386,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5125,
|
| 294 |
+
"grad_norm": 0.012080417945981026,
|
| 295 |
+
"learning_rate": 4.986138888888889e-06,
|
| 296 |
+
"loss": 0.3399,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.525,
|
| 301 |
+
"grad_norm": 0.012398626655340195,
|
| 302 |
+
"learning_rate": 4.9722500000000005e-06,
|
| 303 |
+
"loss": 0.3269,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5375,
|
| 308 |
+
"grad_norm": 0.7642549276351929,
|
| 309 |
+
"learning_rate": 4.958361111111111e-06,
|
| 310 |
+
"loss": 0.2879,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.55,
|
| 315 |
+
"grad_norm": 0.09296104311943054,
|
| 316 |
+
"learning_rate": 4.9444722222222226e-06,
|
| 317 |
+
"loss": 0.3181,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5625,
|
| 322 |
+
"grad_norm": 0.00763101689517498,
|
| 323 |
+
"learning_rate": 4.930583333333334e-06,
|
| 324 |
+
"loss": 0.2816,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.575,
|
| 329 |
+
"grad_norm": 0.01172051951289177,
|
| 330 |
+
"learning_rate": 4.916694444444445e-06,
|
| 331 |
+
"loss": 0.3259,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5875,
|
| 336 |
+
"grad_norm": 0.015954310074448586,
|
| 337 |
+
"learning_rate": 4.902805555555556e-06,
|
| 338 |
+
"loss": 0.2336,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.6,
|
| 343 |
+
"grad_norm": 20.73999786376953,
|
| 344 |
+
"learning_rate": 4.8889166666666674e-06,
|
| 345 |
+
"loss": 0.3255,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.6125,
|
| 350 |
+
"grad_norm": 0.01183092501014471,
|
| 351 |
+
"learning_rate": 4.875027777777778e-06,
|
| 352 |
+
"loss": 0.2953,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.625,
|
| 357 |
+
"grad_norm": 368.5484619140625,
|
| 358 |
+
"learning_rate": 4.8611388888888895e-06,
|
| 359 |
+
"loss": 0.2733,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6375,
|
| 364 |
+
"grad_norm": 0.023890795186161995,
|
| 365 |
+
"learning_rate": 4.847250000000001e-06,
|
| 366 |
+
"loss": 0.2929,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.65,
|
| 371 |
+
"grad_norm": 0.009666858240962029,
|
| 372 |
+
"learning_rate": 4.8333611111111115e-06,
|
| 373 |
+
"loss": 0.2289,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6625,
|
| 378 |
+
"grad_norm": 11.71799373626709,
|
| 379 |
+
"learning_rate": 4.819472222222223e-06,
|
| 380 |
+
"loss": 0.2396,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.675,
|
| 385 |
+
"grad_norm": 173.8921356201172,
|
| 386 |
+
"learning_rate": 4.8055833333333335e-06,
|
| 387 |
+
"loss": 0.2684,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6875,
|
| 392 |
+
"grad_norm": 0.0010307455668225884,
|
| 393 |
+
"learning_rate": 4.791694444444445e-06,
|
| 394 |
+
"loss": 0.2543,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.7,
|
| 399 |
+
"grad_norm": 0.5270601511001587,
|
| 400 |
+
"learning_rate": 4.7778055555555555e-06,
|
| 401 |
+
"loss": 0.2652,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7125,
|
| 406 |
+
"grad_norm": 0.031072473153471947,
|
| 407 |
+
"learning_rate": 4.763916666666667e-06,
|
| 408 |
+
"loss": 0.291,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.725,
|
| 413 |
+
"grad_norm": 0.015740619972348213,
|
| 414 |
+
"learning_rate": 4.750027777777778e-06,
|
| 415 |
+
"loss": 0.2381,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7375,
|
| 420 |
+
"grad_norm": 0.013210597448050976,
|
| 421 |
+
"learning_rate": 4.736138888888889e-06,
|
| 422 |
+
"loss": 0.239,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.75,
|
| 427 |
+
"grad_norm": 0.00735945301130414,
|
| 428 |
+
"learning_rate": 4.72225e-06,
|
| 429 |
+
"loss": 0.1984,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7625,
|
| 434 |
+
"grad_norm": 0.011138584464788437,
|
| 435 |
+
"learning_rate": 4.708361111111112e-06,
|
| 436 |
+
"loss": 0.2401,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.775,
|
| 441 |
+
"grad_norm": 3.5947213172912598,
|
| 442 |
+
"learning_rate": 4.694472222222222e-06,
|
| 443 |
+
"loss": 0.2375,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7875,
|
| 448 |
+
"grad_norm": 0.03575507923960686,
|
| 449 |
+
"learning_rate": 4.680583333333334e-06,
|
| 450 |
+
"loss": 0.264,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.8,
|
| 455 |
+
"grad_norm": 0.011910141445696354,
|
| 456 |
+
"learning_rate": 4.666694444444445e-06,
|
| 457 |
+
"loss": 0.2367,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.8125,
|
| 462 |
+
"grad_norm": 0.015016453340649605,
|
| 463 |
+
"learning_rate": 4.652805555555556e-06,
|
| 464 |
+
"loss": 0.2427,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.825,
|
| 469 |
+
"grad_norm": 129.52890014648438,
|
| 470 |
+
"learning_rate": 4.6389166666666665e-06,
|
| 471 |
+
"loss": 0.2208,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8375,
|
| 476 |
+
"grad_norm": 0.03964228183031082,
|
| 477 |
+
"learning_rate": 4.625027777777778e-06,
|
| 478 |
+
"loss": 0.1863,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.85,
|
| 483 |
+
"grad_norm": 0.06889275461435318,
|
| 484 |
+
"learning_rate": 4.611138888888889e-06,
|
| 485 |
+
"loss": 0.2526,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8625,
|
| 490 |
+
"grad_norm": 0.009284596890211105,
|
| 491 |
+
"learning_rate": 4.59725e-06,
|
| 492 |
+
"loss": 0.2281,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.875,
|
| 497 |
+
"grad_norm": 0.02173932082951069,
|
| 498 |
+
"learning_rate": 4.583361111111111e-06,
|
| 499 |
+
"loss": 0.2357,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8875,
|
| 504 |
+
"grad_norm": 0.2086459994316101,
|
| 505 |
+
"learning_rate": 4.569472222222223e-06,
|
| 506 |
+
"loss": 0.1764,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.9,
|
| 511 |
+
"grad_norm": 0.3403266668319702,
|
| 512 |
+
"learning_rate": 4.555583333333334e-06,
|
| 513 |
+
"loss": 0.2075,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.9125,
|
| 518 |
+
"grad_norm": 0.010942379012703896,
|
| 519 |
+
"learning_rate": 4.541694444444445e-06,
|
| 520 |
+
"loss": 0.2245,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.925,
|
| 525 |
+
"grad_norm": 8.935309597291052e-05,
|
| 526 |
+
"learning_rate": 4.527805555555556e-06,
|
| 527 |
+
"loss": 0.1583,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.9375,
|
| 532 |
+
"grad_norm": 0.0007859256002120674,
|
| 533 |
+
"learning_rate": 4.513916666666668e-06,
|
| 534 |
+
"loss": 0.2016,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.95,
|
| 539 |
+
"grad_norm": 9.582052734913304e-05,
|
| 540 |
+
"learning_rate": 4.500027777777778e-06,
|
| 541 |
+
"loss": 0.2117,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9625,
|
| 546 |
+
"grad_norm": 801.5213012695312,
|
| 547 |
+
"learning_rate": 4.486138888888889e-06,
|
| 548 |
+
"loss": 0.1858,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.975,
|
| 553 |
+
"grad_norm": 0.00019180079107172787,
|
| 554 |
+
"learning_rate": 4.47225e-06,
|
| 555 |
+
"loss": 0.2273,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9875,
|
| 560 |
+
"grad_norm": 0.00854406412690878,
|
| 561 |
+
"learning_rate": 4.458361111111112e-06,
|
| 562 |
+
"loss": 0.2419,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.0,
|
| 567 |
+
"grad_norm": 0.5351490378379822,
|
| 568 |
+
"learning_rate": 4.444472222222222e-06,
|
| 569 |
+
"loss": 0.1964,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.0,
|
| 574 |
+
"eval_accuracy": 0.91205,
|
| 575 |
+
"eval_loss": 0.5335493087768555,
|
| 576 |
+
"eval_runtime": 72.3355,
|
| 577 |
+
"eval_samples_per_second": 276.489,
|
| 578 |
+
"eval_steps_per_second": 69.122,
|
| 579 |
+
"step": 40000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.0125,
|
| 583 |
+
"grad_norm": 0.010720465332269669,
|
| 584 |
+
"learning_rate": 4.430583333333334e-06,
|
| 585 |
+
"loss": 0.1343,
|
| 586 |
+
"step": 40500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.025,
|
| 590 |
+
"grad_norm": 0.001053882297128439,
|
| 591 |
+
"learning_rate": 4.416694444444445e-06,
|
| 592 |
+
"loss": 0.2317,
|
| 593 |
+
"step": 41000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.0375,
|
| 597 |
+
"grad_norm": 0.08478885889053345,
|
| 598 |
+
"learning_rate": 4.402805555555556e-06,
|
| 599 |
+
"loss": 0.1304,
|
| 600 |
+
"step": 41500
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.05,
|
| 604 |
+
"grad_norm": 0.007641951087862253,
|
| 605 |
+
"learning_rate": 4.388916666666667e-06,
|
| 606 |
+
"loss": 0.1912,
|
| 607 |
+
"step": 42000
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 1.0625,
|
| 611 |
+
"grad_norm": 0.0008450501482002437,
|
| 612 |
+
"learning_rate": 4.375027777777779e-06,
|
| 613 |
+
"loss": 0.1816,
|
| 614 |
+
"step": 42500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.075,
|
| 618 |
+
"grad_norm": 0.00660599023103714,
|
| 619 |
+
"learning_rate": 4.361138888888889e-06,
|
| 620 |
+
"loss": 0.1872,
|
| 621 |
+
"step": 43000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.0875,
|
| 625 |
+
"grad_norm": 0.02802414819598198,
|
| 626 |
+
"learning_rate": 4.34725e-06,
|
| 627 |
+
"loss": 0.1797,
|
| 628 |
+
"step": 43500
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.1,
|
| 632 |
+
"grad_norm": 0.04336298257112503,
|
| 633 |
+
"learning_rate": 4.333361111111111e-06,
|
| 634 |
+
"loss": 0.1473,
|
| 635 |
+
"step": 44000
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.1125,
|
| 639 |
+
"grad_norm": 0.0009568212553858757,
|
| 640 |
+
"learning_rate": 4.319472222222223e-06,
|
| 641 |
+
"loss": 0.1572,
|
| 642 |
+
"step": 44500
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.125,
|
| 646 |
+
"grad_norm": 0.001199259888380766,
|
| 647 |
+
"learning_rate": 4.305583333333333e-06,
|
| 648 |
+
"loss": 0.1874,
|
| 649 |
+
"step": 45000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.1375,
|
| 653 |
+
"grad_norm": 0.008190172724425793,
|
| 654 |
+
"learning_rate": 4.291694444444445e-06,
|
| 655 |
+
"loss": 0.2571,
|
| 656 |
+
"step": 45500
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.15,
|
| 660 |
+
"grad_norm": 0.018085084855556488,
|
| 661 |
+
"learning_rate": 4.277805555555556e-06,
|
| 662 |
+
"loss": 0.2197,
|
| 663 |
+
"step": 46000
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.1625,
|
| 667 |
+
"grad_norm": 0.00019737149705179036,
|
| 668 |
+
"learning_rate": 4.263916666666667e-06,
|
| 669 |
+
"loss": 0.1498,
|
| 670 |
+
"step": 46500
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.175,
|
| 674 |
+
"grad_norm": 0.0007391768740490079,
|
| 675 |
+
"learning_rate": 4.250027777777778e-06,
|
| 676 |
+
"loss": 0.1599,
|
| 677 |
+
"step": 47000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.1875,
|
| 681 |
+
"grad_norm": 0.00027270030113868415,
|
| 682 |
+
"learning_rate": 4.2361388888888895e-06,
|
| 683 |
+
"loss": 0.1637,
|
| 684 |
+
"step": 47500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.2,
|
| 688 |
+
"grad_norm": 0.007702063303440809,
|
| 689 |
+
"learning_rate": 4.22225e-06,
|
| 690 |
+
"loss": 0.2234,
|
| 691 |
+
"step": 48000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.2125,
|
| 695 |
+
"grad_norm": 76.7024917602539,
|
| 696 |
+
"learning_rate": 4.2083611111111115e-06,
|
| 697 |
+
"loss": 0.2164,
|
| 698 |
+
"step": 48500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.225,
|
| 702 |
+
"grad_norm": 0.005129904951900244,
|
| 703 |
+
"learning_rate": 4.194472222222223e-06,
|
| 704 |
+
"loss": 0.0964,
|
| 705 |
+
"step": 49000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.2375,
|
| 709 |
+
"grad_norm": 0.005428163334727287,
|
| 710 |
+
"learning_rate": 4.1805833333333336e-06,
|
| 711 |
+
"loss": 0.2027,
|
| 712 |
+
"step": 49500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.25,
|
| 716 |
+
"grad_norm": 0.0023394206073135138,
|
| 717 |
+
"learning_rate": 4.166694444444445e-06,
|
| 718 |
+
"loss": 0.1597,
|
| 719 |
+
"step": 50000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.2625,
|
| 723 |
+
"grad_norm": 3.80388087251049e-06,
|
| 724 |
+
"learning_rate": 4.1528055555555556e-06,
|
| 725 |
+
"loss": 0.1041,
|
| 726 |
+
"step": 50500
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.275,
|
| 730 |
+
"grad_norm": 0.10043651610612869,
|
| 731 |
+
"learning_rate": 4.138916666666667e-06,
|
| 732 |
+
"loss": 0.1689,
|
| 733 |
+
"step": 51000
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 1.2875,
|
| 737 |
+
"grad_norm": 0.8555727601051331,
|
| 738 |
+
"learning_rate": 4.1250277777777784e-06,
|
| 739 |
+
"loss": 0.1732,
|
| 740 |
+
"step": 51500
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.3,
|
| 744 |
+
"grad_norm": 0.036154795438051224,
|
| 745 |
+
"learning_rate": 4.111138888888889e-06,
|
| 746 |
+
"loss": 0.1895,
|
| 747 |
+
"step": 52000
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.3125,
|
| 751 |
+
"grad_norm": 0.12704850733280182,
|
| 752 |
+
"learning_rate": 4.0972500000000005e-06,
|
| 753 |
+
"loss": 0.1476,
|
| 754 |
+
"step": 52500
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.325,
|
| 758 |
+
"grad_norm": 0.037445101886987686,
|
| 759 |
+
"learning_rate": 4.083361111111112e-06,
|
| 760 |
+
"loss": 0.1434,
|
| 761 |
+
"step": 53000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.3375,
|
| 765 |
+
"grad_norm": 1.4244892554415856e-05,
|
| 766 |
+
"learning_rate": 4.0694722222222225e-06,
|
| 767 |
+
"loss": 0.1934,
|
| 768 |
+
"step": 53500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.35,
|
| 772 |
+
"grad_norm": 0.05814299359917641,
|
| 773 |
+
"learning_rate": 4.055583333333334e-06,
|
| 774 |
+
"loss": 0.1674,
|
| 775 |
+
"step": 54000
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 1.3625,
|
| 779 |
+
"grad_norm": 120.2337417602539,
|
| 780 |
+
"learning_rate": 4.041694444444445e-06,
|
| 781 |
+
"loss": 0.155,
|
| 782 |
+
"step": 54500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.375,
|
| 786 |
+
"grad_norm": 0.012538187205791473,
|
| 787 |
+
"learning_rate": 4.027805555555556e-06,
|
| 788 |
+
"loss": 0.1196,
|
| 789 |
+
"step": 55000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1.3875,
|
| 793 |
+
"grad_norm": 0.01885361410677433,
|
| 794 |
+
"learning_rate": 4.0139166666666665e-06,
|
| 795 |
+
"loss": 0.1773,
|
| 796 |
+
"step": 55500
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.4,
|
| 800 |
+
"grad_norm": 0.0020437000785022974,
|
| 801 |
+
"learning_rate": 4.000027777777778e-06,
|
| 802 |
+
"loss": 0.2195,
|
| 803 |
+
"step": 56000
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.4125,
|
| 807 |
+
"grad_norm": 0.0043403045274317265,
|
| 808 |
+
"learning_rate": 3.986138888888889e-06,
|
| 809 |
+
"loss": 0.1506,
|
| 810 |
+
"step": 56500
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.425,
|
| 814 |
+
"grad_norm": 0.029600977897644043,
|
| 815 |
+
"learning_rate": 3.97225e-06,
|
| 816 |
+
"loss": 0.1905,
|
| 817 |
+
"step": 57000
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 1.4375,
|
| 821 |
+
"grad_norm": 0.0036520950961858034,
|
| 822 |
+
"learning_rate": 3.958361111111111e-06,
|
| 823 |
+
"loss": 0.1869,
|
| 824 |
+
"step": 57500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.45,
|
| 828 |
+
"grad_norm": 0.0004638723039533943,
|
| 829 |
+
"learning_rate": 3.944472222222223e-06,
|
| 830 |
+
"loss": 0.1575,
|
| 831 |
+
"step": 58000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 1.4625,
|
| 835 |
+
"grad_norm": 0.00012263878306839615,
|
| 836 |
+
"learning_rate": 3.930583333333333e-06,
|
| 837 |
+
"loss": 0.1624,
|
| 838 |
+
"step": 58500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.475,
|
| 842 |
+
"grad_norm": 132.2960205078125,
|
| 843 |
+
"learning_rate": 3.916694444444445e-06,
|
| 844 |
+
"loss": 0.1547,
|
| 845 |
+
"step": 59000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.4875,
|
| 849 |
+
"grad_norm": 0.002475355053320527,
|
| 850 |
+
"learning_rate": 3.902805555555556e-06,
|
| 851 |
+
"loss": 0.1517,
|
| 852 |
+
"step": 59500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.5,
|
| 856 |
+
"grad_norm": 0.396793931722641,
|
| 857 |
+
"learning_rate": 3.888916666666667e-06,
|
| 858 |
+
"loss": 0.1749,
|
| 859 |
+
"step": 60000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 1.5125,
|
| 863 |
+
"grad_norm": 0.0002843483234755695,
|
| 864 |
+
"learning_rate": 3.8750277777777774e-06,
|
| 865 |
+
"loss": 0.1465,
|
| 866 |
+
"step": 60500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.525,
|
| 870 |
+
"grad_norm": 0.013508698903024197,
|
| 871 |
+
"learning_rate": 3.861138888888889e-06,
|
| 872 |
+
"loss": 0.1794,
|
| 873 |
+
"step": 61000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 1.5375,
|
| 877 |
+
"grad_norm": 0.3459691107273102,
|
| 878 |
+
"learning_rate": 3.84725e-06,
|
| 879 |
+
"loss": 0.1431,
|
| 880 |
+
"step": 61500
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 1.55,
|
| 884 |
+
"grad_norm": 0.0022266607265919447,
|
| 885 |
+
"learning_rate": 3.833361111111111e-06,
|
| 886 |
+
"loss": 0.1725,
|
| 887 |
+
"step": 62000
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 1.5625,
|
| 891 |
+
"grad_norm": 0.015393017791211605,
|
| 892 |
+
"learning_rate": 3.819472222222222e-06,
|
| 893 |
+
"loss": 0.1659,
|
| 894 |
+
"step": 62500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.575,
|
| 898 |
+
"grad_norm": 0.0006895596161484718,
|
| 899 |
+
"learning_rate": 3.8055833333333338e-06,
|
| 900 |
+
"loss": 0.104,
|
| 901 |
+
"step": 63000
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 1.5875,
|
| 905 |
+
"grad_norm": 99.34886932373047,
|
| 906 |
+
"learning_rate": 3.7916944444444448e-06,
|
| 907 |
+
"loss": 0.1676,
|
| 908 |
+
"step": 63500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.6,
|
| 912 |
+
"grad_norm": 0.0647934079170227,
|
| 913 |
+
"learning_rate": 3.7778055555555558e-06,
|
| 914 |
+
"loss": 0.1576,
|
| 915 |
+
"step": 64000
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 1.6125,
|
| 919 |
+
"grad_norm": 0.0019897823221981525,
|
| 920 |
+
"learning_rate": 3.763916666666667e-06,
|
| 921 |
+
"loss": 0.1288,
|
| 922 |
+
"step": 64500
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.625,
|
| 926 |
+
"grad_norm": 0.0011682398617267609,
|
| 927 |
+
"learning_rate": 3.7500277777777782e-06,
|
| 928 |
+
"loss": 0.1508,
|
| 929 |
+
"step": 65000
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.6375,
|
| 933 |
+
"grad_norm": 0.006074314471334219,
|
| 934 |
+
"learning_rate": 3.7361388888888892e-06,
|
| 935 |
+
"loss": 0.1355,
|
| 936 |
+
"step": 65500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.65,
|
| 940 |
+
"grad_norm": 0.0016712377546355128,
|
| 941 |
+
"learning_rate": 3.7222500000000002e-06,
|
| 942 |
+
"loss": 0.1515,
|
| 943 |
+
"step": 66000
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 1.6625,
|
| 947 |
+
"grad_norm": 0.022902656346559525,
|
| 948 |
+
"learning_rate": 3.7083611111111112e-06,
|
| 949 |
+
"loss": 0.1502,
|
| 950 |
+
"step": 66500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.675,
|
| 954 |
+
"grad_norm": 1.1224092304473743e-05,
|
| 955 |
+
"learning_rate": 3.6944722222222223e-06,
|
| 956 |
+
"loss": 0.1649,
|
| 957 |
+
"step": 67000
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.6875,
|
| 961 |
+
"grad_norm": 0.00015980980242602527,
|
| 962 |
+
"learning_rate": 3.6805833333333337e-06,
|
| 963 |
+
"loss": 0.184,
|
| 964 |
+
"step": 67500
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.7,
|
| 968 |
+
"grad_norm": 0.008271472528576851,
|
| 969 |
+
"learning_rate": 3.6666944444444447e-06,
|
| 970 |
+
"loss": 0.1036,
|
| 971 |
+
"step": 68000
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.7125,
|
| 975 |
+
"grad_norm": 0.003849911969155073,
|
| 976 |
+
"learning_rate": 3.6528055555555557e-06,
|
| 977 |
+
"loss": 0.1225,
|
| 978 |
+
"step": 68500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.725,
|
| 982 |
+
"grad_norm": 0.00014347155229188502,
|
| 983 |
+
"learning_rate": 3.638916666666667e-06,
|
| 984 |
+
"loss": 0.1213,
|
| 985 |
+
"step": 69000
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.7375,
|
| 989 |
+
"grad_norm": 0.0031831359956413507,
|
| 990 |
+
"learning_rate": 3.625027777777778e-06,
|
| 991 |
+
"loss": 0.1658,
|
| 992 |
+
"step": 69500
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.75,
|
| 996 |
+
"grad_norm": 0.00018772439216263592,
|
| 997 |
+
"learning_rate": 3.611138888888889e-06,
|
| 998 |
+
"loss": 0.1193,
|
| 999 |
+
"step": 70000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.7625,
|
| 1003 |
+
"grad_norm": 0.0017813814338296652,
|
| 1004 |
+
"learning_rate": 3.5972500000000006e-06,
|
| 1005 |
+
"loss": 0.1326,
|
| 1006 |
+
"step": 70500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.775,
|
| 1010 |
+
"grad_norm": 0.006025127600878477,
|
| 1011 |
+
"learning_rate": 3.5833611111111116e-06,
|
| 1012 |
+
"loss": 0.0978,
|
| 1013 |
+
"step": 71000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.7875,
|
| 1017 |
+
"grad_norm": 0.10651105642318726,
|
| 1018 |
+
"learning_rate": 3.5694722222222226e-06,
|
| 1019 |
+
"loss": 0.163,
|
| 1020 |
+
"step": 71500
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.8,
|
| 1024 |
+
"grad_norm": 0.0005099499830976129,
|
| 1025 |
+
"learning_rate": 3.555583333333333e-06,
|
| 1026 |
+
"loss": 0.1218,
|
| 1027 |
+
"step": 72000
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.8125,
|
| 1031 |
+
"grad_norm": 0.013583320192992687,
|
| 1032 |
+
"learning_rate": 3.5416944444444446e-06,
|
| 1033 |
+
"loss": 0.0974,
|
| 1034 |
+
"step": 72500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.825,
|
| 1038 |
+
"grad_norm": 0.004824354313313961,
|
| 1039 |
+
"learning_rate": 3.5278055555555556e-06,
|
| 1040 |
+
"loss": 0.1885,
|
| 1041 |
+
"step": 73000
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.8375,
|
| 1045 |
+
"grad_norm": 0.017673935741186142,
|
| 1046 |
+
"learning_rate": 3.513916666666667e-06,
|
| 1047 |
+
"loss": 0.1381,
|
| 1048 |
+
"step": 73500
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.85,
|
| 1052 |
+
"grad_norm": 3.5670293527800823e-06,
|
| 1053 |
+
"learning_rate": 3.500027777777778e-06,
|
| 1054 |
+
"loss": 0.1369,
|
| 1055 |
+
"step": 74000
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.8625,
|
| 1059 |
+
"grad_norm": 0.03615639731287956,
|
| 1060 |
+
"learning_rate": 3.486138888888889e-06,
|
| 1061 |
+
"loss": 0.1453,
|
| 1062 |
+
"step": 74500
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.875,
|
| 1066 |
+
"grad_norm": 0.00019953282026108354,
|
| 1067 |
+
"learning_rate": 3.4722500000000005e-06,
|
| 1068 |
+
"loss": 0.117,
|
| 1069 |
+
"step": 75000
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 1.8875,
|
| 1073 |
+
"grad_norm": 9.114891952322068e-08,
|
| 1074 |
+
"learning_rate": 3.4583611111111115e-06,
|
| 1075 |
+
"loss": 0.1227,
|
| 1076 |
+
"step": 75500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.9,
|
| 1080 |
+
"grad_norm": 0.0070320055820047855,
|
| 1081 |
+
"learning_rate": 3.4444722222222225e-06,
|
| 1082 |
+
"loss": 0.0892,
|
| 1083 |
+
"step": 76000
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.9125,
|
| 1087 |
+
"grad_norm": 2.0117835447308607e-05,
|
| 1088 |
+
"learning_rate": 3.430583333333334e-06,
|
| 1089 |
+
"loss": 0.1272,
|
| 1090 |
+
"step": 76500
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.925,
|
| 1094 |
+
"grad_norm": 1.597428672539536e-05,
|
| 1095 |
+
"learning_rate": 3.416694444444445e-06,
|
| 1096 |
+
"loss": 0.1555,
|
| 1097 |
+
"step": 77000
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.9375,
|
| 1101 |
+
"grad_norm": 0.03601458668708801,
|
| 1102 |
+
"learning_rate": 3.4028055555555556e-06,
|
| 1103 |
+
"loss": 0.1518,
|
| 1104 |
+
"step": 77500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.95,
|
| 1108 |
+
"grad_norm": 68.35214233398438,
|
| 1109 |
+
"learning_rate": 3.3889166666666666e-06,
|
| 1110 |
+
"loss": 0.1481,
|
| 1111 |
+
"step": 78000
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.9625,
|
| 1115 |
+
"grad_norm": 0.03208393603563309,
|
| 1116 |
+
"learning_rate": 3.375027777777778e-06,
|
| 1117 |
+
"loss": 0.1816,
|
| 1118 |
+
"step": 78500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.975,
|
| 1122 |
+
"grad_norm": 0.04399827867746353,
|
| 1123 |
+
"learning_rate": 3.361138888888889e-06,
|
| 1124 |
+
"loss": 0.149,
|
| 1125 |
+
"step": 79000
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.9875,
|
| 1129 |
+
"grad_norm": 0.007236841134727001,
|
| 1130 |
+
"learning_rate": 3.34725e-06,
|
| 1131 |
+
"loss": 0.1421,
|
| 1132 |
+
"step": 79500
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.0,
|
| 1136 |
+
"grad_norm": 475.6263427734375,
|
| 1137 |
+
"learning_rate": 3.3333611111111114e-06,
|
| 1138 |
+
"loss": 0.1498,
|
| 1139 |
+
"step": 80000
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.0,
|
| 1143 |
+
"eval_accuracy": 0.98015,
|
| 1144 |
+
"eval_loss": 0.11060585081577301,
|
| 1145 |
+
"eval_runtime": 71.1478,
|
| 1146 |
+
"eval_samples_per_second": 281.105,
|
| 1147 |
+
"eval_steps_per_second": 70.276,
|
| 1148 |
+
"step": 80000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.0125,
|
| 1152 |
+
"grad_norm": 8.577239896112587e-06,
|
| 1153 |
+
"learning_rate": 3.3194722222222225e-06,
|
| 1154 |
+
"loss": 0.0677,
|
| 1155 |
+
"step": 80500
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 2.025,
|
| 1159 |
+
"grad_norm": 2.4226756067946553e-07,
|
| 1160 |
+
"learning_rate": 3.305583333333334e-06,
|
| 1161 |
+
"loss": 0.0873,
|
| 1162 |
+
"step": 81000
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 2.0375,
|
| 1166 |
+
"grad_norm": 0.0019143010722473264,
|
| 1167 |
+
"learning_rate": 3.291694444444445e-06,
|
| 1168 |
+
"loss": 0.1247,
|
| 1169 |
+
"step": 81500
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 2.05,
|
| 1173 |
+
"grad_norm": 0.00012104820052627474,
|
| 1174 |
+
"learning_rate": 3.277805555555556e-06,
|
| 1175 |
+
"loss": 0.0585,
|
| 1176 |
+
"step": 82000
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.0625,
|
| 1180 |
+
"grad_norm": 0.02980652078986168,
|
| 1181 |
+
"learning_rate": 3.2639166666666673e-06,
|
| 1182 |
+
"loss": 0.0722,
|
| 1183 |
+
"step": 82500
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 2.075,
|
| 1187 |
+
"grad_norm": 0.00023673578107263893,
|
| 1188 |
+
"learning_rate": 3.250027777777778e-06,
|
| 1189 |
+
"loss": 0.1211,
|
| 1190 |
+
"step": 83000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.0875,
|
| 1194 |
+
"grad_norm": 0.020454034209251404,
|
| 1195 |
+
"learning_rate": 3.236138888888889e-06,
|
| 1196 |
+
"loss": 0.0758,
|
| 1197 |
+
"step": 83500
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 2.1,
|
| 1201 |
+
"grad_norm": 2.698595835681772e-06,
|
| 1202 |
+
"learning_rate": 3.22225e-06,
|
| 1203 |
+
"loss": 0.0527,
|
| 1204 |
+
"step": 84000
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 2.1125,
|
| 1208 |
+
"grad_norm": 0.0003798000980168581,
|
| 1209 |
+
"learning_rate": 3.2083611111111114e-06,
|
| 1210 |
+
"loss": 0.1335,
|
| 1211 |
+
"step": 84500
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 2.125,
|
| 1215 |
+
"grad_norm": 0.0007345640915445983,
|
| 1216 |
+
"learning_rate": 3.1944722222222224e-06,
|
| 1217 |
+
"loss": 0.1113,
|
| 1218 |
+
"step": 85000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 2.1375,
|
| 1222 |
+
"grad_norm": 0.09974172711372375,
|
| 1223 |
+
"learning_rate": 3.1805833333333334e-06,
|
| 1224 |
+
"loss": 0.0601,
|
| 1225 |
+
"step": 85500
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 2.15,
|
| 1229 |
+
"grad_norm": 0.018733147531747818,
|
| 1230 |
+
"learning_rate": 3.166694444444445e-06,
|
| 1231 |
+
"loss": 0.1287,
|
| 1232 |
+
"step": 86000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.1625,
|
| 1236 |
+
"grad_norm": 0.00047316361451521516,
|
| 1237 |
+
"learning_rate": 3.152805555555556e-06,
|
| 1238 |
+
"loss": 0.0931,
|
| 1239 |
+
"step": 86500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 2.175,
|
| 1243 |
+
"grad_norm": 0.00016918570327106863,
|
| 1244 |
+
"learning_rate": 3.1389166666666673e-06,
|
| 1245 |
+
"loss": 0.0967,
|
| 1246 |
+
"step": 87000
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 2.1875,
|
| 1250 |
+
"grad_norm": 1.8574954083305784e-05,
|
| 1251 |
+
"learning_rate": 3.1250277777777783e-06,
|
| 1252 |
+
"loss": 0.0994,
|
| 1253 |
+
"step": 87500
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 2.2,
|
| 1257 |
+
"grad_norm": 9.292660251958296e-05,
|
| 1258 |
+
"learning_rate": 3.1111388888888893e-06,
|
| 1259 |
+
"loss": 0.0941,
|
| 1260 |
+
"step": 88000
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 2.2125,
|
| 1264 |
+
"grad_norm": 2.387504309808719e-06,
|
| 1265 |
+
"learning_rate": 3.09725e-06,
|
| 1266 |
+
"loss": 0.0806,
|
| 1267 |
+
"step": 88500
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 2.225,
|
| 1271 |
+
"grad_norm": 0.011588960886001587,
|
| 1272 |
+
"learning_rate": 3.0833611111111113e-06,
|
| 1273 |
+
"loss": 0.1497,
|
| 1274 |
+
"step": 89000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.2375,
|
| 1278 |
+
"grad_norm": 0.007856378331780434,
|
| 1279 |
+
"learning_rate": 3.0694722222222223e-06,
|
| 1280 |
+
"loss": 0.1076,
|
| 1281 |
+
"step": 89500
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 2.25,
|
| 1285 |
+
"grad_norm": 0.014046729542315006,
|
| 1286 |
+
"learning_rate": 3.0555833333333333e-06,
|
| 1287 |
+
"loss": 0.0754,
|
| 1288 |
+
"step": 90000
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 2.2625,
|
| 1292 |
+
"grad_norm": 0.13570299744606018,
|
| 1293 |
+
"learning_rate": 3.0416944444444447e-06,
|
| 1294 |
+
"loss": 0.0647,
|
| 1295 |
+
"step": 90500
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 2.275,
|
| 1299 |
+
"grad_norm": 1.2789449783667806e-06,
|
| 1300 |
+
"learning_rate": 3.0278055555555558e-06,
|
| 1301 |
+
"loss": 0.0433,
|
| 1302 |
+
"step": 91000
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 2.2875,
|
| 1306 |
+
"grad_norm": 0.00010895993909798563,
|
| 1307 |
+
"learning_rate": 3.0139166666666668e-06,
|
| 1308 |
+
"loss": 0.0798,
|
| 1309 |
+
"step": 91500
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 2.3,
|
| 1313 |
+
"grad_norm": 0.0030339015647768974,
|
| 1314 |
+
"learning_rate": 3.000027777777778e-06,
|
| 1315 |
+
"loss": 0.1252,
|
| 1316 |
+
"step": 92000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.3125,
|
| 1320 |
+
"grad_norm": 0.12522538006305695,
|
| 1321 |
+
"learning_rate": 2.986138888888889e-06,
|
| 1322 |
+
"loss": 0.0911,
|
| 1323 |
+
"step": 92500
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 2.325,
|
| 1327 |
+
"grad_norm": 0.00940883718430996,
|
| 1328 |
+
"learning_rate": 2.9722500000000002e-06,
|
| 1329 |
+
"loss": 0.0999,
|
| 1330 |
+
"step": 93000
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 2.3375,
|
| 1334 |
+
"grad_norm": 0.05087483301758766,
|
| 1335 |
+
"learning_rate": 2.9583611111111116e-06,
|
| 1336 |
+
"loss": 0.1069,
|
| 1337 |
+
"step": 93500
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 2.35,
|
| 1341 |
+
"grad_norm": 0.0046332115307450294,
|
| 1342 |
+
"learning_rate": 2.9444722222222227e-06,
|
| 1343 |
+
"loss": 0.0814,
|
| 1344 |
+
"step": 94000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 2.3625,
|
| 1348 |
+
"grad_norm": 4.976536729373038e-05,
|
| 1349 |
+
"learning_rate": 2.9305833333333332e-06,
|
| 1350 |
+
"loss": 0.0749,
|
| 1351 |
+
"step": 94500
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 2.375,
|
| 1355 |
+
"grad_norm": 0.13259227573871613,
|
| 1356 |
+
"learning_rate": 2.9166944444444447e-06,
|
| 1357 |
+
"loss": 0.0826,
|
| 1358 |
+
"step": 95000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 2.3875,
|
| 1362 |
+
"grad_norm": 8.53817709867144e-07,
|
| 1363 |
+
"learning_rate": 2.9028055555555557e-06,
|
| 1364 |
+
"loss": 0.0924,
|
| 1365 |
+
"step": 95500
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 2.4,
|
| 1369 |
+
"grad_norm": 0.017462341114878654,
|
| 1370 |
+
"learning_rate": 2.8889166666666667e-06,
|
| 1371 |
+
"loss": 0.0861,
|
| 1372 |
+
"step": 96000
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.4125,
|
| 1376 |
+
"grad_norm": 0.010675615631043911,
|
| 1377 |
+
"learning_rate": 2.875027777777778e-06,
|
| 1378 |
+
"loss": 0.1213,
|
| 1379 |
+
"step": 96500
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 2.425,
|
| 1383 |
+
"grad_norm": 0.0014171106740832329,
|
| 1384 |
+
"learning_rate": 2.861138888888889e-06,
|
| 1385 |
+
"loss": 0.0655,
|
| 1386 |
+
"step": 97000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 2.4375,
|
| 1390 |
+
"grad_norm": 1.936705729121968e-07,
|
| 1391 |
+
"learning_rate": 2.84725e-06,
|
| 1392 |
+
"loss": 0.1056,
|
| 1393 |
+
"step": 97500
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 2.45,
|
| 1397 |
+
"grad_norm": 3.3484088817203883e-06,
|
| 1398 |
+
"learning_rate": 2.8333611111111116e-06,
|
| 1399 |
+
"loss": 0.0603,
|
| 1400 |
+
"step": 98000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 2.4625,
|
| 1404 |
+
"grad_norm": 0.025326106697320938,
|
| 1405 |
+
"learning_rate": 2.8194722222222226e-06,
|
| 1406 |
+
"loss": 0.0809,
|
| 1407 |
+
"step": 98500
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 2.475,
|
| 1411 |
+
"grad_norm": 0.00743254367262125,
|
| 1412 |
+
"learning_rate": 2.8055833333333336e-06,
|
| 1413 |
+
"loss": 0.1357,
|
| 1414 |
+
"step": 99000
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.4875,
|
| 1418 |
+
"grad_norm": 0.01988009549677372,
|
| 1419 |
+
"learning_rate": 2.791694444444445e-06,
|
| 1420 |
+
"loss": 0.0974,
|
| 1421 |
+
"step": 99500
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 2.5,
|
| 1425 |
+
"grad_norm": 0.0004283917951397598,
|
| 1426 |
+
"learning_rate": 2.7778055555555556e-06,
|
| 1427 |
+
"loss": 0.1095,
|
| 1428 |
+
"step": 100000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 2.5125,
|
| 1432 |
+
"grad_norm": 0.005049830302596092,
|
| 1433 |
+
"learning_rate": 2.7639166666666666e-06,
|
| 1434 |
+
"loss": 0.1019,
|
| 1435 |
+
"step": 100500
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 2.525,
|
| 1439 |
+
"grad_norm": 2.7473144655232318e-05,
|
| 1440 |
+
"learning_rate": 2.750027777777778e-06,
|
| 1441 |
+
"loss": 0.0916,
|
| 1442 |
+
"step": 101000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 2.5375,
|
| 1446 |
+
"grad_norm": 0.19764241576194763,
|
| 1447 |
+
"learning_rate": 2.736138888888889e-06,
|
| 1448 |
+
"loss": 0.1209,
|
| 1449 |
+
"step": 101500
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 2.55,
|
| 1453 |
+
"grad_norm": 0.15940068662166595,
|
| 1454 |
+
"learning_rate": 2.72225e-06,
|
| 1455 |
+
"loss": 0.039,
|
| 1456 |
+
"step": 102000
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.5625,
|
| 1460 |
+
"grad_norm": 0.030904898419976234,
|
| 1461 |
+
"learning_rate": 2.7083611111111115e-06,
|
| 1462 |
+
"loss": 0.0581,
|
| 1463 |
+
"step": 102500
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 2.575,
|
| 1467 |
+
"grad_norm": 0.5066580176353455,
|
| 1468 |
+
"learning_rate": 2.6944722222222225e-06,
|
| 1469 |
+
"loss": 0.0971,
|
| 1470 |
+
"step": 103000
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.5875,
|
| 1474 |
+
"grad_norm": 2.1701941577134676e-08,
|
| 1475 |
+
"learning_rate": 2.6805833333333335e-06,
|
| 1476 |
+
"loss": 0.0984,
|
| 1477 |
+
"step": 103500
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 2.6,
|
| 1481 |
+
"grad_norm": 0.0060079507529735565,
|
| 1482 |
+
"learning_rate": 2.666694444444445e-06,
|
| 1483 |
+
"loss": 0.1176,
|
| 1484 |
+
"step": 104000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 2.6125,
|
| 1488 |
+
"grad_norm": 0.01262228935956955,
|
| 1489 |
+
"learning_rate": 2.652805555555556e-06,
|
| 1490 |
+
"loss": 0.0999,
|
| 1491 |
+
"step": 104500
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 2.625,
|
| 1495 |
+
"grad_norm": 0.004787284880876541,
|
| 1496 |
+
"learning_rate": 2.638916666666667e-06,
|
| 1497 |
+
"loss": 0.1031,
|
| 1498 |
+
"step": 105000
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 2.6375,
|
| 1502 |
+
"grad_norm": 0.004941838793456554,
|
| 1503 |
+
"learning_rate": 2.6250277777777776e-06,
|
| 1504 |
+
"loss": 0.1018,
|
| 1505 |
+
"step": 105500
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 2.65,
|
| 1509 |
+
"grad_norm": 2.6329735192121007e-06,
|
| 1510 |
+
"learning_rate": 2.611138888888889e-06,
|
| 1511 |
+
"loss": 0.1081,
|
| 1512 |
+
"step": 106000
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.6625,
|
| 1516 |
+
"grad_norm": 0.028844570741057396,
|
| 1517 |
+
"learning_rate": 2.59725e-06,
|
| 1518 |
+
"loss": 0.0558,
|
| 1519 |
+
"step": 106500
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 2.675,
|
| 1523 |
+
"grad_norm": 1.3074130666979045e-09,
|
| 1524 |
+
"learning_rate": 2.583361111111111e-06,
|
| 1525 |
+
"loss": 0.098,
|
| 1526 |
+
"step": 107000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 2.6875,
|
| 1530 |
+
"grad_norm": 0.1365622878074646,
|
| 1531 |
+
"learning_rate": 2.5694722222222224e-06,
|
| 1532 |
+
"loss": 0.0967,
|
| 1533 |
+
"step": 107500
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 2.7,
|
| 1537 |
+
"grad_norm": 1.7111767647293163e-07,
|
| 1538 |
+
"learning_rate": 2.5555833333333334e-06,
|
| 1539 |
+
"loss": 0.0829,
|
| 1540 |
+
"step": 108000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 2.7125,
|
| 1544 |
+
"grad_norm": 1.4454585652856622e-05,
|
| 1545 |
+
"learning_rate": 2.541694444444445e-06,
|
| 1546 |
+
"loss": 0.1038,
|
| 1547 |
+
"step": 108500
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 2.725,
|
| 1551 |
+
"grad_norm": 0.3521536588668823,
|
| 1552 |
+
"learning_rate": 2.527805555555556e-06,
|
| 1553 |
+
"loss": 0.0797,
|
| 1554 |
+
"step": 109000
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 2.7375,
|
| 1558 |
+
"grad_norm": 0.00021774623019155115,
|
| 1559 |
+
"learning_rate": 2.513916666666667e-06,
|
| 1560 |
+
"loss": 0.1158,
|
| 1561 |
+
"step": 109500
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 2.75,
|
| 1565 |
+
"grad_norm": 0.0001049950733431615,
|
| 1566 |
+
"learning_rate": 2.5000277777777783e-06,
|
| 1567 |
+
"loss": 0.0928,
|
| 1568 |
+
"step": 110000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 2.7625,
|
| 1572 |
+
"grad_norm": 3.3473178007170645e-08,
|
| 1573 |
+
"learning_rate": 2.486138888888889e-06,
|
| 1574 |
+
"loss": 0.0979,
|
| 1575 |
+
"step": 110500
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 2.775,
|
| 1579 |
+
"grad_norm": 0.003986213356256485,
|
| 1580 |
+
"learning_rate": 2.4722500000000003e-06,
|
| 1581 |
+
"loss": 0.0958,
|
| 1582 |
+
"step": 111000
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 2.7875,
|
| 1586 |
+
"grad_norm": 5.052464530308498e-06,
|
| 1587 |
+
"learning_rate": 2.4583611111111114e-06,
|
| 1588 |
+
"loss": 0.0445,
|
| 1589 |
+
"step": 111500
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 2.8,
|
| 1593 |
+
"grad_norm": 1.3031937839969032e-07,
|
| 1594 |
+
"learning_rate": 2.4444722222222224e-06,
|
| 1595 |
+
"loss": 0.0673,
|
| 1596 |
+
"step": 112000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 2.8125,
|
| 1600 |
+
"grad_norm": 0.013562874868512154,
|
| 1601 |
+
"learning_rate": 2.4305833333333334e-06,
|
| 1602 |
+
"loss": 0.0778,
|
| 1603 |
+
"step": 112500
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 2.825,
|
| 1607 |
+
"grad_norm": 1.478714057157049e-05,
|
| 1608 |
+
"learning_rate": 2.4166944444444444e-06,
|
| 1609 |
+
"loss": 0.0978,
|
| 1610 |
+
"step": 113000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 2.8375,
|
| 1614 |
+
"grad_norm": 0.15072928369045258,
|
| 1615 |
+
"learning_rate": 2.402805555555556e-06,
|
| 1616 |
+
"loss": 0.036,
|
| 1617 |
+
"step": 113500
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 2.85,
|
| 1621 |
+
"grad_norm": 0.15145017206668854,
|
| 1622 |
+
"learning_rate": 2.388916666666667e-06,
|
| 1623 |
+
"loss": 0.0779,
|
| 1624 |
+
"step": 114000
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 2.8625,
|
| 1628 |
+
"grad_norm": 0.0001674382365308702,
|
| 1629 |
+
"learning_rate": 2.3750277777777783e-06,
|
| 1630 |
+
"loss": 0.1006,
|
| 1631 |
+
"step": 114500
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 2.875,
|
| 1635 |
+
"grad_norm": 0.0002086051827063784,
|
| 1636 |
+
"learning_rate": 2.3611388888888893e-06,
|
| 1637 |
+
"loss": 0.0734,
|
| 1638 |
+
"step": 115000
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 2.8875,
|
| 1642 |
+
"grad_norm": 164.32354736328125,
|
| 1643 |
+
"learning_rate": 2.3472500000000003e-06,
|
| 1644 |
+
"loss": 0.0729,
|
| 1645 |
+
"step": 115500
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 2.9,
|
| 1649 |
+
"grad_norm": 0.012316230684518814,
|
| 1650 |
+
"learning_rate": 2.3333611111111113e-06,
|
| 1651 |
+
"loss": 0.0984,
|
| 1652 |
+
"step": 116000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 2.9125,
|
| 1656 |
+
"grad_norm": 0.32554274797439575,
|
| 1657 |
+
"learning_rate": 2.3194722222222223e-06,
|
| 1658 |
+
"loss": 0.0906,
|
| 1659 |
+
"step": 116500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 2.925,
|
| 1663 |
+
"grad_norm": 7.992193218342436e-07,
|
| 1664 |
+
"learning_rate": 2.3055833333333337e-06,
|
| 1665 |
+
"loss": 0.0891,
|
| 1666 |
+
"step": 117000
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 2.9375,
|
| 1670 |
+
"grad_norm": 0.001029053470119834,
|
| 1671 |
+
"learning_rate": 2.2916944444444447e-06,
|
| 1672 |
+
"loss": 0.104,
|
| 1673 |
+
"step": 117500
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 2.95,
|
| 1677 |
+
"grad_norm": 0.0013586468994617462,
|
| 1678 |
+
"learning_rate": 2.2778055555555557e-06,
|
| 1679 |
+
"loss": 0.04,
|
| 1680 |
+
"step": 118000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 2.9625,
|
| 1684 |
+
"grad_norm": 5.0012647989206016e-05,
|
| 1685 |
+
"learning_rate": 2.2639166666666667e-06,
|
| 1686 |
+
"loss": 0.0689,
|
| 1687 |
+
"step": 118500
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 2.975,
|
| 1691 |
+
"grad_norm": 0.045165080577135086,
|
| 1692 |
+
"learning_rate": 2.2500277777777778e-06,
|
| 1693 |
+
"loss": 0.0758,
|
| 1694 |
+
"step": 119000
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 2.9875,
|
| 1698 |
+
"grad_norm": 4.196267582301516e-07,
|
| 1699 |
+
"learning_rate": 2.236138888888889e-06,
|
| 1700 |
+
"loss": 0.0596,
|
| 1701 |
+
"step": 119500
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 3.0,
|
| 1705 |
+
"grad_norm": 0.00039908240432851017,
|
| 1706 |
+
"learning_rate": 2.22225e-06,
|
| 1707 |
+
"loss": 0.0775,
|
| 1708 |
+
"step": 120000
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 3.0,
|
| 1712 |
+
"eval_accuracy": 0.9856,
|
| 1713 |
+
"eval_loss": 0.09523609280586243,
|
| 1714 |
+
"eval_runtime": 70.8442,
|
| 1715 |
+
"eval_samples_per_second": 282.31,
|
| 1716 |
+
"eval_steps_per_second": 70.577,
|
| 1717 |
+
"step": 120000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.0125,
|
| 1721 |
+
"grad_norm": 0.0008493961067870259,
|
| 1722 |
+
"learning_rate": 2.208361111111111e-06,
|
| 1723 |
+
"loss": 0.0756,
|
| 1724 |
+
"step": 120500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.025,
|
| 1728 |
+
"grad_norm": 3.1318024412030354e-05,
|
| 1729 |
+
"learning_rate": 2.1944722222222222e-06,
|
| 1730 |
+
"loss": 0.061,
|
| 1731 |
+
"step": 121000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.0375,
|
| 1735 |
+
"grad_norm": 7.05946590073836e-08,
|
| 1736 |
+
"learning_rate": 2.1805833333333336e-06,
|
| 1737 |
+
"loss": 0.032,
|
| 1738 |
+
"step": 121500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.05,
|
| 1742 |
+
"grad_norm": 2.9236124987619405e-07,
|
| 1743 |
+
"learning_rate": 2.1666944444444447e-06,
|
| 1744 |
+
"loss": 0.049,
|
| 1745 |
+
"step": 122000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.0625,
|
| 1749 |
+
"grad_norm": 3.1436852623301093e-06,
|
| 1750 |
+
"learning_rate": 2.1528055555555557e-06,
|
| 1751 |
+
"loss": 0.0465,
|
| 1752 |
+
"step": 122500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.075,
|
| 1756 |
+
"grad_norm": 7.781285660257709e-08,
|
| 1757 |
+
"learning_rate": 2.138916666666667e-06,
|
| 1758 |
+
"loss": 0.0702,
|
| 1759 |
+
"step": 123000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.0875,
|
| 1763 |
+
"grad_norm": 2.085931782858097e-07,
|
| 1764 |
+
"learning_rate": 2.125027777777778e-06,
|
| 1765 |
+
"loss": 0.0448,
|
| 1766 |
+
"step": 123500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 3.1,
|
| 1770 |
+
"grad_norm": 1024.1812744140625,
|
| 1771 |
+
"learning_rate": 2.111138888888889e-06,
|
| 1772 |
+
"loss": 0.0898,
|
| 1773 |
+
"step": 124000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 3.1125,
|
| 1777 |
+
"grad_norm": 1.351660188220194e-07,
|
| 1778 |
+
"learning_rate": 2.09725e-06,
|
| 1779 |
+
"loss": 0.0431,
|
| 1780 |
+
"step": 124500
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 3.125,
|
| 1784 |
+
"grad_norm": 0.001149096991866827,
|
| 1785 |
+
"learning_rate": 2.083361111111111e-06,
|
| 1786 |
+
"loss": 0.0486,
|
| 1787 |
+
"step": 125000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 3.1375,
|
| 1791 |
+
"grad_norm": 2.472369851602707e-05,
|
| 1792 |
+
"learning_rate": 2.0694722222222226e-06,
|
| 1793 |
+
"loss": 0.0539,
|
| 1794 |
+
"step": 125500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 3.15,
|
| 1798 |
+
"grad_norm": 6.923924615875876e-07,
|
| 1799 |
+
"learning_rate": 2.0555833333333336e-06,
|
| 1800 |
+
"loss": 0.0751,
|
| 1801 |
+
"step": 126000
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 3.1625,
|
| 1805 |
+
"grad_norm": 3.3076943850574025e-08,
|
| 1806 |
+
"learning_rate": 2.0416944444444446e-06,
|
| 1807 |
+
"loss": 0.0216,
|
| 1808 |
+
"step": 126500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 3.175,
|
| 1812 |
+
"grad_norm": 1.5224570404370752e-07,
|
| 1813 |
+
"learning_rate": 2.0278055555555556e-06,
|
| 1814 |
+
"loss": 0.0425,
|
| 1815 |
+
"step": 127000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 3.1875,
|
| 1819 |
+
"grad_norm": 2.331158430024516e-05,
|
| 1820 |
+
"learning_rate": 2.0139166666666666e-06,
|
| 1821 |
+
"loss": 0.0568,
|
| 1822 |
+
"step": 127500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 3.2,
|
| 1826 |
+
"grad_norm": 0.0012923574540764093,
|
| 1827 |
+
"learning_rate": 2.000027777777778e-06,
|
| 1828 |
+
"loss": 0.0467,
|
| 1829 |
+
"step": 128000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 3.2125,
|
| 1833 |
+
"grad_norm": 0.0013359450967982411,
|
| 1834 |
+
"learning_rate": 1.986138888888889e-06,
|
| 1835 |
+
"loss": 0.0781,
|
| 1836 |
+
"step": 128500
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 3.225,
|
| 1840 |
+
"grad_norm": 0.007183945272117853,
|
| 1841 |
+
"learning_rate": 1.9722500000000005e-06,
|
| 1842 |
+
"loss": 0.0328,
|
| 1843 |
+
"step": 129000
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 3.2375,
|
| 1847 |
+
"grad_norm": 2.5895735234371386e-05,
|
| 1848 |
+
"learning_rate": 1.958361111111111e-06,
|
| 1849 |
+
"loss": 0.0565,
|
| 1850 |
+
"step": 129500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 3.25,
|
| 1854 |
+
"grad_norm": 0.003281608922407031,
|
| 1855 |
+
"learning_rate": 1.9444722222222225e-06,
|
| 1856 |
+
"loss": 0.0573,
|
| 1857 |
+
"step": 130000
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 3.2625,
|
| 1861 |
+
"grad_norm": 7.837111581920908e-08,
|
| 1862 |
+
"learning_rate": 1.9305833333333335e-06,
|
| 1863 |
+
"loss": 0.0129,
|
| 1864 |
+
"step": 130500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 3.275,
|
| 1868 |
+
"grad_norm": 1.010946107271593e-06,
|
| 1869 |
+
"learning_rate": 1.9166944444444445e-06,
|
| 1870 |
+
"loss": 0.0716,
|
| 1871 |
+
"step": 131000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 3.2875,
|
| 1875 |
+
"grad_norm": 1.7364520317642018e-05,
|
| 1876 |
+
"learning_rate": 1.9028055555555557e-06,
|
| 1877 |
+
"loss": 0.0342,
|
| 1878 |
+
"step": 131500
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 3.3,
|
| 1882 |
+
"grad_norm": 2.711913194008986e-10,
|
| 1883 |
+
"learning_rate": 1.888916666666667e-06,
|
| 1884 |
+
"loss": 0.0652,
|
| 1885 |
+
"step": 132000
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 3.3125,
|
| 1889 |
+
"grad_norm": 2.3724902348476462e-05,
|
| 1890 |
+
"learning_rate": 1.8750277777777777e-06,
|
| 1891 |
+
"loss": 0.0512,
|
| 1892 |
+
"step": 132500
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 3.325,
|
| 1896 |
+
"grad_norm": 1.995879850280602e-10,
|
| 1897 |
+
"learning_rate": 1.861138888888889e-06,
|
| 1898 |
+
"loss": 0.051,
|
| 1899 |
+
"step": 133000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 3.3375,
|
| 1903 |
+
"grad_norm": 1.3715705238226406e-09,
|
| 1904 |
+
"learning_rate": 1.8472500000000002e-06,
|
| 1905 |
+
"loss": 0.0173,
|
| 1906 |
+
"step": 133500
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 3.35,
|
| 1910 |
+
"grad_norm": 2.5874431486272442e-08,
|
| 1911 |
+
"learning_rate": 1.8333611111111114e-06,
|
| 1912 |
+
"loss": 0.0265,
|
| 1913 |
+
"step": 134000
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 3.3625,
|
| 1917 |
+
"grad_norm": 9.791951924853493e-07,
|
| 1918 |
+
"learning_rate": 1.8194722222222224e-06,
|
| 1919 |
+
"loss": 0.0639,
|
| 1920 |
+
"step": 134500
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 3.375,
|
| 1924 |
+
"grad_norm": 1.8817563826600825e-11,
|
| 1925 |
+
"learning_rate": 1.8055833333333334e-06,
|
| 1926 |
+
"loss": 0.0213,
|
| 1927 |
+
"step": 135000
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 3.3875,
|
| 1931 |
+
"grad_norm": 1.4598784900954342e-06,
|
| 1932 |
+
"learning_rate": 1.7916944444444444e-06,
|
| 1933 |
+
"loss": 0.0356,
|
| 1934 |
+
"step": 135500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 3.4,
|
| 1938 |
+
"grad_norm": 4.592724280882976e-07,
|
| 1939 |
+
"learning_rate": 1.7778055555555557e-06,
|
| 1940 |
+
"loss": 0.0767,
|
| 1941 |
+
"step": 136000
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 3.4125,
|
| 1945 |
+
"grad_norm": 1.918931502586929e-07,
|
| 1946 |
+
"learning_rate": 1.7639166666666669e-06,
|
| 1947 |
+
"loss": 0.0157,
|
| 1948 |
+
"step": 136500
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 3.425,
|
| 1952 |
+
"grad_norm": 5.836361651745392e-07,
|
| 1953 |
+
"learning_rate": 1.750027777777778e-06,
|
| 1954 |
+
"loss": 0.0424,
|
| 1955 |
+
"step": 137000
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 3.4375,
|
| 1959 |
+
"grad_norm": 5.298739324643975e-06,
|
| 1960 |
+
"learning_rate": 1.736138888888889e-06,
|
| 1961 |
+
"loss": 0.0655,
|
| 1962 |
+
"step": 137500
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 3.45,
|
| 1966 |
+
"grad_norm": 7.815018761903048e-05,
|
| 1967 |
+
"learning_rate": 1.7222500000000001e-06,
|
| 1968 |
+
"loss": 0.0194,
|
| 1969 |
+
"step": 138000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.4625,
|
| 1973 |
+
"grad_norm": 0.0004915116005577147,
|
| 1974 |
+
"learning_rate": 1.7083611111111111e-06,
|
| 1975 |
+
"loss": 0.0949,
|
| 1976 |
+
"step": 138500
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.475,
|
| 1980 |
+
"grad_norm": 0.002174359979107976,
|
| 1981 |
+
"learning_rate": 1.6944722222222223e-06,
|
| 1982 |
+
"loss": 0.0587,
|
| 1983 |
+
"step": 139000
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.4875,
|
| 1987 |
+
"grad_norm": 0.0017873262986540794,
|
| 1988 |
+
"learning_rate": 1.6805833333333336e-06,
|
| 1989 |
+
"loss": 0.0659,
|
| 1990 |
+
"step": 139500
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.5,
|
| 1994 |
+
"grad_norm": 0.00011011360766133294,
|
| 1995 |
+
"learning_rate": 1.6666944444444446e-06,
|
| 1996 |
+
"loss": 0.0454,
|
| 1997 |
+
"step": 140000
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.5125,
|
| 2001 |
+
"grad_norm": 5.8530247770249844e-05,
|
| 2002 |
+
"learning_rate": 1.6528055555555556e-06,
|
| 2003 |
+
"loss": 0.0209,
|
| 2004 |
+
"step": 140500
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.525,
|
| 2008 |
+
"grad_norm": 0.0005035810172557831,
|
| 2009 |
+
"learning_rate": 1.6389166666666668e-06,
|
| 2010 |
+
"loss": 0.0575,
|
| 2011 |
+
"step": 141000
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.5375,
|
| 2015 |
+
"grad_norm": 7.48745515011251e-05,
|
| 2016 |
+
"learning_rate": 1.6250277777777778e-06,
|
| 2017 |
+
"loss": 0.0354,
|
| 2018 |
+
"step": 141500
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.55,
|
| 2022 |
+
"grad_norm": 2.5532403014949523e-05,
|
| 2023 |
+
"learning_rate": 1.611138888888889e-06,
|
| 2024 |
+
"loss": 0.0304,
|
| 2025 |
+
"step": 142000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.5625,
|
| 2029 |
+
"grad_norm": 8.711179543752223e-05,
|
| 2030 |
+
"learning_rate": 1.5972500000000003e-06,
|
| 2031 |
+
"loss": 0.0483,
|
| 2032 |
+
"step": 142500
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.575,
|
| 2036 |
+
"grad_norm": 1.006107552825597e-08,
|
| 2037 |
+
"learning_rate": 1.5833611111111113e-06,
|
| 2038 |
+
"loss": 0.04,
|
| 2039 |
+
"step": 143000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.5875,
|
| 2043 |
+
"grad_norm": 0.00023778372269589454,
|
| 2044 |
+
"learning_rate": 1.5694722222222223e-06,
|
| 2045 |
+
"loss": 0.0633,
|
| 2046 |
+
"step": 143500
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.6,
|
| 2050 |
+
"grad_norm": 0.09549811482429504,
|
| 2051 |
+
"learning_rate": 1.5555833333333335e-06,
|
| 2052 |
+
"loss": 0.0593,
|
| 2053 |
+
"step": 144000
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.6125,
|
| 2057 |
+
"grad_norm": 0.0007943428936414421,
|
| 2058 |
+
"learning_rate": 1.5416944444444445e-06,
|
| 2059 |
+
"loss": 0.0664,
|
| 2060 |
+
"step": 144500
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.625,
|
| 2064 |
+
"grad_norm": 187.2425994873047,
|
| 2065 |
+
"learning_rate": 1.5278055555555557e-06,
|
| 2066 |
+
"loss": 0.0574,
|
| 2067 |
+
"step": 145000
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.6375,
|
| 2071 |
+
"grad_norm": 8.670787792652845e-06,
|
| 2072 |
+
"learning_rate": 1.513916666666667e-06,
|
| 2073 |
+
"loss": 0.0284,
|
| 2074 |
+
"step": 145500
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.65,
|
| 2078 |
+
"grad_norm": 6.026363553246483e-05,
|
| 2079 |
+
"learning_rate": 1.500027777777778e-06,
|
| 2080 |
+
"loss": 0.0188,
|
| 2081 |
+
"step": 146000
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.6625,
|
| 2085 |
+
"grad_norm": 2.375714514002425e-09,
|
| 2086 |
+
"learning_rate": 1.486138888888889e-06,
|
| 2087 |
+
"loss": 0.0257,
|
| 2088 |
+
"step": 146500
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.675,
|
| 2092 |
+
"grad_norm": 6.045565532986075e-05,
|
| 2093 |
+
"learning_rate": 1.47225e-06,
|
| 2094 |
+
"loss": 0.065,
|
| 2095 |
+
"step": 147000
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.6875,
|
| 2099 |
+
"grad_norm": 0.0009056538692675531,
|
| 2100 |
+
"learning_rate": 1.4583611111111112e-06,
|
| 2101 |
+
"loss": 0.0666,
|
| 2102 |
+
"step": 147500
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.7,
|
| 2106 |
+
"grad_norm": 0.013074205256998539,
|
| 2107 |
+
"learning_rate": 1.4444722222222224e-06,
|
| 2108 |
+
"loss": 0.0439,
|
| 2109 |
+
"step": 148000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.7125,
|
| 2113 |
+
"grad_norm": 0.0008992800139822066,
|
| 2114 |
+
"learning_rate": 1.4305833333333336e-06,
|
| 2115 |
+
"loss": 0.0542,
|
| 2116 |
+
"step": 148500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.725,
|
| 2120 |
+
"grad_norm": 1.5302573430631128e-08,
|
| 2121 |
+
"learning_rate": 1.4166944444444444e-06,
|
| 2122 |
+
"loss": 0.059,
|
| 2123 |
+
"step": 149000
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.7375,
|
| 2127 |
+
"grad_norm": 0.015462064184248447,
|
| 2128 |
+
"learning_rate": 1.4028055555555556e-06,
|
| 2129 |
+
"loss": 0.0444,
|
| 2130 |
+
"step": 149500
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.75,
|
| 2134 |
+
"grad_norm": 2.3089251044439152e-05,
|
| 2135 |
+
"learning_rate": 1.3889166666666667e-06,
|
| 2136 |
+
"loss": 0.0488,
|
| 2137 |
+
"step": 150000
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.7625,
|
| 2141 |
+
"grad_norm": 0.029342608526349068,
|
| 2142 |
+
"learning_rate": 1.3750277777777779e-06,
|
| 2143 |
+
"loss": 0.0387,
|
| 2144 |
+
"step": 150500
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.775,
|
| 2148 |
+
"grad_norm": 0.16643966734409332,
|
| 2149 |
+
"learning_rate": 1.361138888888889e-06,
|
| 2150 |
+
"loss": 0.0496,
|
| 2151 |
+
"step": 151000
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.7875,
|
| 2155 |
+
"grad_norm": 3.40620975869399e-10,
|
| 2156 |
+
"learning_rate": 1.3472500000000003e-06,
|
| 2157 |
+
"loss": 0.037,
|
| 2158 |
+
"step": 151500
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.8,
|
| 2162 |
+
"grad_norm": 1.0341510980893531e-09,
|
| 2163 |
+
"learning_rate": 1.3333611111111111e-06,
|
| 2164 |
+
"loss": 0.0434,
|
| 2165 |
+
"step": 152000
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.8125,
|
| 2169 |
+
"grad_norm": 6.396485696313903e-05,
|
| 2170 |
+
"learning_rate": 1.3194722222222223e-06,
|
| 2171 |
+
"loss": 0.0924,
|
| 2172 |
+
"step": 152500
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.825,
|
| 2176 |
+
"grad_norm": 2.9705923831890857e-10,
|
| 2177 |
+
"learning_rate": 1.3055833333333333e-06,
|
| 2178 |
+
"loss": 0.0319,
|
| 2179 |
+
"step": 153000
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.8375,
|
| 2183 |
+
"grad_norm": 0.00013349021901376545,
|
| 2184 |
+
"learning_rate": 1.2916944444444446e-06,
|
| 2185 |
+
"loss": 0.028,
|
| 2186 |
+
"step": 153500
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.85,
|
| 2190 |
+
"grad_norm": 1.4059290620593856e-08,
|
| 2191 |
+
"learning_rate": 1.2778055555555558e-06,
|
| 2192 |
+
"loss": 0.0219,
|
| 2193 |
+
"step": 154000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 3.8625,
|
| 2197 |
+
"grad_norm": 1.5280780871762545e-06,
|
| 2198 |
+
"learning_rate": 1.263916666666667e-06,
|
| 2199 |
+
"loss": 0.0522,
|
| 2200 |
+
"step": 154500
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 3.875,
|
| 2204 |
+
"grad_norm": 4.033211264964365e-10,
|
| 2205 |
+
"learning_rate": 1.2500277777777778e-06,
|
| 2206 |
+
"loss": 0.0331,
|
| 2207 |
+
"step": 155000
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 3.8875,
|
| 2211 |
+
"grad_norm": 0.004559781868010759,
|
| 2212 |
+
"learning_rate": 1.236138888888889e-06,
|
| 2213 |
+
"loss": 0.0628,
|
| 2214 |
+
"step": 155500
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 3.9,
|
| 2218 |
+
"grad_norm": 4.622693813871592e-05,
|
| 2219 |
+
"learning_rate": 1.22225e-06,
|
| 2220 |
+
"loss": 0.0114,
|
| 2221 |
+
"step": 156000
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 3.9125,
|
| 2225 |
+
"grad_norm": 3.810692703609675e-07,
|
| 2226 |
+
"learning_rate": 1.2083611111111112e-06,
|
| 2227 |
+
"loss": 0.0466,
|
| 2228 |
+
"step": 156500
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 3.925,
|
| 2232 |
+
"grad_norm": 9.439718269277364e-07,
|
| 2233 |
+
"learning_rate": 1.1944722222222223e-06,
|
| 2234 |
+
"loss": 0.0058,
|
| 2235 |
+
"step": 157000
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 3.9375,
|
| 2239 |
+
"grad_norm": 2.499190010496477e-08,
|
| 2240 |
+
"learning_rate": 1.1805833333333335e-06,
|
| 2241 |
+
"loss": 0.0573,
|
| 2242 |
+
"step": 157500
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 3.95,
|
| 2246 |
+
"grad_norm": 2.330422574914337e-07,
|
| 2247 |
+
"learning_rate": 1.1666944444444447e-06,
|
| 2248 |
+
"loss": 0.0372,
|
| 2249 |
+
"step": 158000
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 3.9625,
|
| 2253 |
+
"grad_norm": 2.389850806139293e-08,
|
| 2254 |
+
"learning_rate": 1.1528055555555557e-06,
|
| 2255 |
+
"loss": 0.0124,
|
| 2256 |
+
"step": 158500
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.975,
|
| 2260 |
+
"grad_norm": 1.5795259544759688e-09,
|
| 2261 |
+
"learning_rate": 1.1389166666666667e-06,
|
| 2262 |
+
"loss": 0.0521,
|
| 2263 |
+
"step": 159000
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 3.9875,
|
| 2267 |
+
"grad_norm": 1.8120088762785258e-09,
|
| 2268 |
+
"learning_rate": 1.1250277777777777e-06,
|
| 2269 |
+
"loss": 0.0346,
|
| 2270 |
+
"step": 159500
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 4.0,
|
| 2274 |
+
"grad_norm": 7.511826161099577e-10,
|
| 2275 |
+
"learning_rate": 1.111138888888889e-06,
|
| 2276 |
+
"loss": 0.0671,
|
| 2277 |
+
"step": 160000
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 4.0,
|
| 2281 |
+
"eval_accuracy": 0.9812,
|
| 2282 |
+
"eval_loss": 0.14172975718975067,
|
| 2283 |
+
"eval_runtime": 71.1854,
|
| 2284 |
+
"eval_samples_per_second": 280.957,
|
| 2285 |
+
"eval_steps_per_second": 70.239,
|
| 2286 |
+
"step": 160000
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 4.0125,
|
| 2290 |
+
"grad_norm": 1.181604005751069e-07,
|
| 2291 |
+
"learning_rate": 1.0972500000000002e-06,
|
| 2292 |
+
"loss": 0.0149,
|
| 2293 |
+
"step": 160500
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 4.025,
|
| 2297 |
+
"grad_norm": 4.751647429657169e-05,
|
| 2298 |
+
"learning_rate": 1.0833611111111112e-06,
|
| 2299 |
+
"loss": 0.0229,
|
| 2300 |
+
"step": 161000
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 4.0375,
|
| 2304 |
+
"grad_norm": 1.4241804180414874e-08,
|
| 2305 |
+
"learning_rate": 1.0694722222222224e-06,
|
| 2306 |
+
"loss": 0.0083,
|
| 2307 |
+
"step": 161500
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 4.05,
|
| 2311 |
+
"grad_norm": 2.3136502136367199e-07,
|
| 2312 |
+
"learning_rate": 1.0555833333333334e-06,
|
| 2313 |
+
"loss": 0.0282,
|
| 2314 |
+
"step": 162000
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 4.0625,
|
| 2318 |
+
"grad_norm": 3.991391167801339e-06,
|
| 2319 |
+
"learning_rate": 1.0416944444444444e-06,
|
| 2320 |
+
"loss": 0.0386,
|
| 2321 |
+
"step": 162500
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 4.075,
|
| 2325 |
+
"grad_norm": 1.6578444217429933e-07,
|
| 2326 |
+
"learning_rate": 1.0278055555555556e-06,
|
| 2327 |
+
"loss": 0.0116,
|
| 2328 |
+
"step": 163000
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 4.0875,
|
| 2332 |
+
"grad_norm": 2.1836729501956142e-05,
|
| 2333 |
+
"learning_rate": 1.0139166666666669e-06,
|
| 2334 |
+
"loss": 0.0008,
|
| 2335 |
+
"step": 163500
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 4.1,
|
| 2339 |
+
"grad_norm": 3.056627804198797e-07,
|
| 2340 |
+
"learning_rate": 1.0000277777777779e-06,
|
| 2341 |
+
"loss": 0.0081,
|
| 2342 |
+
"step": 164000
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 4.1125,
|
| 2346 |
+
"grad_norm": 0.000962151272688061,
|
| 2347 |
+
"learning_rate": 9.86138888888889e-07,
|
| 2348 |
+
"loss": 0.0166,
|
| 2349 |
+
"step": 164500
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 4.125,
|
| 2353 |
+
"grad_norm": 4.749813042881357e-11,
|
| 2354 |
+
"learning_rate": 9.7225e-07,
|
| 2355 |
+
"loss": 0.0086,
|
| 2356 |
+
"step": 165000
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 4.1375,
|
| 2360 |
+
"grad_norm": 7.004540503174894e-09,
|
| 2361 |
+
"learning_rate": 9.58361111111111e-07,
|
| 2362 |
+
"loss": 0.0103,
|
| 2363 |
+
"step": 165500
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 4.15,
|
| 2367 |
+
"grad_norm": 0.0017288046656176448,
|
| 2368 |
+
"learning_rate": 9.444722222222223e-07,
|
| 2369 |
+
"loss": 0.0356,
|
| 2370 |
+
"step": 166000
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 4.1625,
|
| 2374 |
+
"grad_norm": 8.125256023561178e-09,
|
| 2375 |
+
"learning_rate": 9.305833333333334e-07,
|
| 2376 |
+
"loss": 0.021,
|
| 2377 |
+
"step": 166500
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 4.175,
|
| 2381 |
+
"grad_norm": 8.191600642248886e-08,
|
| 2382 |
+
"learning_rate": 9.166944444444444e-07,
|
| 2383 |
+
"loss": 0.0111,
|
| 2384 |
+
"step": 167000
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 4.1875,
|
| 2388 |
+
"grad_norm": 1.6114168888137215e-09,
|
| 2389 |
+
"learning_rate": 9.028055555555557e-07,
|
| 2390 |
+
"loss": 0.0002,
|
| 2391 |
+
"step": 167500
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 4.2,
|
| 2395 |
+
"grad_norm": 2.477648308740754e-07,
|
| 2396 |
+
"learning_rate": 8.889166666666667e-07,
|
| 2397 |
+
"loss": 0.012,
|
| 2398 |
+
"step": 168000
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 4.2125,
|
| 2402 |
+
"grad_norm": 7.130810899980133e-06,
|
| 2403 |
+
"learning_rate": 8.750277777777778e-07,
|
| 2404 |
+
"loss": 0.0303,
|
| 2405 |
+
"step": 168500
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 4.225,
|
| 2409 |
+
"grad_norm": 1.779966041226544e-08,
|
| 2410 |
+
"learning_rate": 8.61138888888889e-07,
|
| 2411 |
+
"loss": 0.013,
|
| 2412 |
+
"step": 169000
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 4.2375,
|
| 2416 |
+
"grad_norm": 0.004164457321166992,
|
| 2417 |
+
"learning_rate": 8.4725e-07,
|
| 2418 |
+
"loss": 0.0281,
|
| 2419 |
+
"step": 169500
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 4.25,
|
| 2423 |
+
"grad_norm": 7.877612451245852e-10,
|
| 2424 |
+
"learning_rate": 8.333611111111111e-07,
|
| 2425 |
+
"loss": 0.028,
|
| 2426 |
+
"step": 170000
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 4.2625,
|
| 2430 |
+
"grad_norm": 1.2292772133548624e-08,
|
| 2431 |
+
"learning_rate": 8.194722222222224e-07,
|
| 2432 |
+
"loss": 0.0217,
|
| 2433 |
+
"step": 170500
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 4.275,
|
| 2437 |
+
"grad_norm": 4.138699105649124e-11,
|
| 2438 |
+
"learning_rate": 8.055833333333334e-07,
|
| 2439 |
+
"loss": 0.0236,
|
| 2440 |
+
"step": 171000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 4.2875,
|
| 2444 |
+
"grad_norm": 2.8552069579745876e-06,
|
| 2445 |
+
"learning_rate": 7.916944444444445e-07,
|
| 2446 |
+
"loss": 0.0266,
|
| 2447 |
+
"step": 171500
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 4.3,
|
| 2451 |
+
"grad_norm": 4.926955111256293e-12,
|
| 2452 |
+
"learning_rate": 7.778055555555557e-07,
|
| 2453 |
+
"loss": 0.018,
|
| 2454 |
+
"step": 172000
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 4.3125,
|
| 2458 |
+
"grad_norm": 2.0034612191466294e-08,
|
| 2459 |
+
"learning_rate": 7.639166666666667e-07,
|
| 2460 |
+
"loss": 0.0284,
|
| 2461 |
+
"step": 172500
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 4.325,
|
| 2465 |
+
"grad_norm": 0.00030996339046396315,
|
| 2466 |
+
"learning_rate": 7.500277777777778e-07,
|
| 2467 |
+
"loss": 0.0122,
|
| 2468 |
+
"step": 173000
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 4.3375,
|
| 2472 |
+
"grad_norm": 1.602331940375734e-05,
|
| 2473 |
+
"learning_rate": 7.36138888888889e-07,
|
| 2474 |
+
"loss": 0.0294,
|
| 2475 |
+
"step": 173500
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 4.35,
|
| 2479 |
+
"grad_norm": 6.264875196393405e-08,
|
| 2480 |
+
"learning_rate": 7.2225e-07,
|
| 2481 |
+
"loss": 0.0297,
|
| 2482 |
+
"step": 174000
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 4.3625,
|
| 2486 |
+
"grad_norm": 2.1285784441715805e-06,
|
| 2487 |
+
"learning_rate": 7.083611111111112e-07,
|
| 2488 |
+
"loss": 0.0,
|
| 2489 |
+
"step": 174500
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 4.375,
|
| 2493 |
+
"grad_norm": 1.7763302821549587e-05,
|
| 2494 |
+
"learning_rate": 6.944722222222222e-07,
|
| 2495 |
+
"loss": 0.0219,
|
| 2496 |
+
"step": 175000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 4.3875,
|
| 2500 |
+
"grad_norm": 1.3622946326563579e-08,
|
| 2501 |
+
"learning_rate": 6.805833333333334e-07,
|
| 2502 |
+
"loss": 0.0222,
|
| 2503 |
+
"step": 175500
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 4.4,
|
| 2507 |
+
"grad_norm": 2.147145465869471e-07,
|
| 2508 |
+
"learning_rate": 6.666944444444445e-07,
|
| 2509 |
+
"loss": 0.021,
|
| 2510 |
+
"step": 176000
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 4.4125,
|
| 2514 |
+
"grad_norm": 1.905289437331703e-08,
|
| 2515 |
+
"learning_rate": 6.528055555555555e-07,
|
| 2516 |
+
"loss": 0.0082,
|
| 2517 |
+
"step": 176500
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 4.425,
|
| 2521 |
+
"grad_norm": 2.2939485380679692e-11,
|
| 2522 |
+
"learning_rate": 6.389166666666667e-07,
|
| 2523 |
+
"loss": 0.0133,
|
| 2524 |
+
"step": 177000
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 4.4375,
|
| 2528 |
+
"grad_norm": 1.8313154214411043e-05,
|
| 2529 |
+
"learning_rate": 6.250277777777779e-07,
|
| 2530 |
+
"loss": 0.0133,
|
| 2531 |
+
"step": 177500
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 4.45,
|
| 2535 |
+
"grad_norm": 0.0007046947139315307,
|
| 2536 |
+
"learning_rate": 6.11138888888889e-07,
|
| 2537 |
+
"loss": 0.0297,
|
| 2538 |
+
"step": 178000
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 4.4625,
|
| 2542 |
+
"grad_norm": 5.863841465725272e-07,
|
| 2543 |
+
"learning_rate": 5.972500000000001e-07,
|
| 2544 |
+
"loss": 0.0054,
|
| 2545 |
+
"step": 178500
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 4.475,
|
| 2549 |
+
"grad_norm": 6.64712516140753e-08,
|
| 2550 |
+
"learning_rate": 5.833611111111111e-07,
|
| 2551 |
+
"loss": 0.0282,
|
| 2552 |
+
"step": 179000
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 4.4875,
|
| 2556 |
+
"grad_norm": 2.658403097655082e-08,
|
| 2557 |
+
"learning_rate": 5.694722222222223e-07,
|
| 2558 |
+
"loss": 0.0164,
|
| 2559 |
+
"step": 179500
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 4.5,
|
| 2563 |
+
"grad_norm": 1.954035013795874e-07,
|
| 2564 |
+
"learning_rate": 5.555833333333334e-07,
|
| 2565 |
+
"loss": 0.0184,
|
| 2566 |
+
"step": 180000
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 4.5125,
|
| 2570 |
+
"grad_norm": 1.2052239414805932e-11,
|
| 2571 |
+
"learning_rate": 5.416944444444444e-07,
|
| 2572 |
+
"loss": 0.0182,
|
| 2573 |
+
"step": 180500
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 4.525,
|
| 2577 |
+
"grad_norm": 1.1565572499705468e-08,
|
| 2578 |
+
"learning_rate": 5.278055555555556e-07,
|
| 2579 |
+
"loss": 0.0137,
|
| 2580 |
+
"step": 181000
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 4.5375,
|
| 2584 |
+
"grad_norm": 9.083069585358317e-08,
|
| 2585 |
+
"learning_rate": 5.139166666666668e-07,
|
| 2586 |
+
"loss": 0.0056,
|
| 2587 |
+
"step": 181500
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 4.55,
|
| 2591 |
+
"grad_norm": 4.157520681891125e-12,
|
| 2592 |
+
"learning_rate": 5.000277777777778e-07,
|
| 2593 |
+
"loss": 0.019,
|
| 2594 |
+
"step": 182000
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 4.5625,
|
| 2598 |
+
"grad_norm": 7.844753781682812e-06,
|
| 2599 |
+
"learning_rate": 4.861388888888889e-07,
|
| 2600 |
+
"loss": 0.0059,
|
| 2601 |
+
"step": 182500
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 4.575,
|
| 2605 |
+
"grad_norm": 2.1387011273077405e-08,
|
| 2606 |
+
"learning_rate": 4.7225000000000006e-07,
|
| 2607 |
+
"loss": 0.0132,
|
| 2608 |
+
"step": 183000
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 4.5875,
|
| 2612 |
+
"grad_norm": 3.293918027225118e-08,
|
| 2613 |
+
"learning_rate": 4.583611111111112e-07,
|
| 2614 |
+
"loss": 0.0402,
|
| 2615 |
+
"step": 183500
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 4.6,
|
| 2619 |
+
"grad_norm": 3.6600565636035753e-06,
|
| 2620 |
+
"learning_rate": 4.4447222222222224e-07,
|
| 2621 |
+
"loss": 0.0207,
|
| 2622 |
+
"step": 184000
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 4.6125,
|
| 2626 |
+
"grad_norm": 9.770146425580606e-05,
|
| 2627 |
+
"learning_rate": 4.3058333333333335e-07,
|
| 2628 |
+
"loss": 0.0076,
|
| 2629 |
+
"step": 184500
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 4.625,
|
| 2633 |
+
"grad_norm": 7.285264587153506e-07,
|
| 2634 |
+
"learning_rate": 4.166944444444445e-07,
|
| 2635 |
+
"loss": 0.0218,
|
| 2636 |
+
"step": 185000
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 4.6375,
|
| 2640 |
+
"grad_norm": 3.197037426461158e-11,
|
| 2641 |
+
"learning_rate": 4.028055555555556e-07,
|
| 2642 |
+
"loss": 0.0002,
|
| 2643 |
+
"step": 185500
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 4.65,
|
| 2647 |
+
"grad_norm": 1.8801161161263735e-07,
|
| 2648 |
+
"learning_rate": 3.889166666666667e-07,
|
| 2649 |
+
"loss": 0.0001,
|
| 2650 |
+
"step": 186000
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 4.6625,
|
| 2654 |
+
"grad_norm": 5.221526766518991e-09,
|
| 2655 |
+
"learning_rate": 3.750277777777778e-07,
|
| 2656 |
+
"loss": 0.0,
|
| 2657 |
+
"step": 186500
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 4.675,
|
| 2661 |
+
"grad_norm": 1.6803708424717456e-10,
|
| 2662 |
+
"learning_rate": 3.6113888888888893e-07,
|
| 2663 |
+
"loss": 0.0065,
|
| 2664 |
+
"step": 187000
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 4.6875,
|
| 2668 |
+
"grad_norm": 0.00025249188183806837,
|
| 2669 |
+
"learning_rate": 3.4725000000000004e-07,
|
| 2670 |
+
"loss": 0.0198,
|
| 2671 |
+
"step": 187500
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 4.7,
|
| 2675 |
+
"grad_norm": 6.064842850150853e-09,
|
| 2676 |
+
"learning_rate": 3.333611111111111e-07,
|
| 2677 |
+
"loss": 0.0076,
|
| 2678 |
+
"step": 188000
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 4.7125,
|
| 2682 |
+
"grad_norm": 1.0817297280141247e-08,
|
| 2683 |
+
"learning_rate": 3.1947222222222227e-07,
|
| 2684 |
+
"loss": 0.0255,
|
| 2685 |
+
"step": 188500
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 4.725,
|
| 2689 |
+
"grad_norm": 3.801495118693765e-09,
|
| 2690 |
+
"learning_rate": 3.0558333333333333e-07,
|
| 2691 |
+
"loss": 0.0118,
|
| 2692 |
+
"step": 189000
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 4.7375,
|
| 2696 |
+
"grad_norm": 1.5158904947271123e-12,
|
| 2697 |
+
"learning_rate": 2.916944444444445e-07,
|
| 2698 |
+
"loss": 0.0066,
|
| 2699 |
+
"step": 189500
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 4.75,
|
| 2703 |
+
"grad_norm": 3.0599636602346436e-08,
|
| 2704 |
+
"learning_rate": 2.7780555555555556e-07,
|
| 2705 |
+
"loss": 0.0126,
|
| 2706 |
+
"step": 190000
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 4.7625,
|
| 2710 |
+
"grad_norm": 0.0005276433657854795,
|
| 2711 |
+
"learning_rate": 2.639166666666667e-07,
|
| 2712 |
+
"loss": 0.0084,
|
| 2713 |
+
"step": 190500
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 4.775,
|
| 2717 |
+
"grad_norm": 4.551525023543945e-08,
|
| 2718 |
+
"learning_rate": 2.500277777777778e-07,
|
| 2719 |
+
"loss": 0.0,
|
| 2720 |
+
"step": 191000
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 4.7875,
|
| 2724 |
+
"grad_norm": 5.455982332591702e-08,
|
| 2725 |
+
"learning_rate": 2.361388888888889e-07,
|
| 2726 |
+
"loss": 0.0113,
|
| 2727 |
+
"step": 191500
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 4.8,
|
| 2731 |
+
"grad_norm": 1.5818011888768524e-05,
|
| 2732 |
+
"learning_rate": 2.2225000000000005e-07,
|
| 2733 |
+
"loss": 0.0213,
|
| 2734 |
+
"step": 192000
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 4.8125,
|
| 2738 |
+
"grad_norm": 5.3856580052524805e-05,
|
| 2739 |
+
"learning_rate": 2.0836111111111114e-07,
|
| 2740 |
+
"loss": 0.0112,
|
| 2741 |
+
"step": 192500
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 4.825,
|
| 2745 |
+
"grad_norm": 0.0007711928919889033,
|
| 2746 |
+
"learning_rate": 1.9447222222222225e-07,
|
| 2747 |
+
"loss": 0.0,
|
| 2748 |
+
"step": 193000
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 4.8375,
|
| 2752 |
+
"grad_norm": 240.61659240722656,
|
| 2753 |
+
"learning_rate": 1.8058333333333334e-07,
|
| 2754 |
+
"loss": 0.0078,
|
| 2755 |
+
"step": 193500
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 4.85,
|
| 2759 |
+
"grad_norm": 5.4270851701510026e-11,
|
| 2760 |
+
"learning_rate": 1.6669444444444445e-07,
|
| 2761 |
+
"loss": 0.017,
|
| 2762 |
+
"step": 194000
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 4.8625,
|
| 2766 |
+
"grad_norm": 2.3522264314124186e-07,
|
| 2767 |
+
"learning_rate": 1.5280555555555557e-07,
|
| 2768 |
+
"loss": 0.0113,
|
| 2769 |
+
"step": 194500
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 4.875,
|
| 2773 |
+
"grad_norm": 4.281631404978725e-08,
|
| 2774 |
+
"learning_rate": 1.3891666666666668e-07,
|
| 2775 |
+
"loss": 0.0066,
|
| 2776 |
+
"step": 195000
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 4.8875,
|
| 2780 |
+
"grad_norm": 2.0812564116567955e-07,
|
| 2781 |
+
"learning_rate": 1.250277777777778e-07,
|
| 2782 |
+
"loss": 0.0088,
|
| 2783 |
+
"step": 195500
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 4.9,
|
| 2787 |
+
"grad_norm": 1.5096416561277692e-08,
|
| 2788 |
+
"learning_rate": 1.111388888888889e-07,
|
| 2789 |
+
"loss": 0.0135,
|
| 2790 |
+
"step": 196000
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 4.9125,
|
| 2794 |
+
"grad_norm": 1.7817910702433437e-08,
|
| 2795 |
+
"learning_rate": 9.725e-08,
|
| 2796 |
+
"loss": 0.006,
|
| 2797 |
+
"step": 196500
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 4.925,
|
| 2801 |
+
"grad_norm": 1.9435563558545255e-07,
|
| 2802 |
+
"learning_rate": 8.336111111111113e-08,
|
| 2803 |
+
"loss": 0.0,
|
| 2804 |
+
"step": 197000
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 4.9375,
|
| 2808 |
+
"grad_norm": 2.460150483329926e-08,
|
| 2809 |
+
"learning_rate": 6.947222222222223e-08,
|
| 2810 |
+
"loss": 0.0094,
|
| 2811 |
+
"step": 197500
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 4.95,
|
| 2815 |
+
"grad_norm": 3.353856072862982e-07,
|
| 2816 |
+
"learning_rate": 5.558333333333334e-08,
|
| 2817 |
+
"loss": 0.0191,
|
| 2818 |
+
"step": 198000
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 4.9625,
|
| 2822 |
+
"grad_norm": 0.0009476316627115011,
|
| 2823 |
+
"learning_rate": 4.1694444444444446e-08,
|
| 2824 |
+
"loss": 0.0158,
|
| 2825 |
+
"step": 198500
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 4.975,
|
| 2829 |
+
"grad_norm": 7.387540676973003e-07,
|
| 2830 |
+
"learning_rate": 2.7805555555555557e-08,
|
| 2831 |
+
"loss": 0.0,
|
| 2832 |
+
"step": 199000
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 4.9875,
|
| 2836 |
+
"grad_norm": 1.0438948372382129e-08,
|
| 2837 |
+
"learning_rate": 1.3916666666666669e-08,
|
| 2838 |
+
"loss": 0.0278,
|
| 2839 |
+
"step": 199500
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 5.0,
|
| 2843 |
+
"grad_norm": 6.0067055496215715e-12,
|
| 2844 |
+
"learning_rate": 2.777777777777778e-11,
|
| 2845 |
+
"loss": 0.0027,
|
| 2846 |
+
"step": 200000
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 5.0,
|
| 2850 |
+
"eval_accuracy": 0.98965,
|
| 2851 |
+
"eval_loss": 0.10565917938947678,
|
| 2852 |
+
"eval_runtime": 71.164,
|
| 2853 |
+
"eval_samples_per_second": 281.041,
|
| 2854 |
+
"eval_steps_per_second": 70.26,
|
| 2855 |
+
"step": 200000
|
| 2856 |
+
}
|
| 2857 |
+
],
|
| 2858 |
+
"logging_steps": 500,
|
| 2859 |
+
"max_steps": 200000,
|
| 2860 |
+
"num_input_tokens_seen": 0,
|
| 2861 |
+
"num_train_epochs": 5,
|
| 2862 |
+
"save_steps": 500,
|
| 2863 |
+
"stateful_callbacks": {
|
| 2864 |
+
"TrainerControl": {
|
| 2865 |
+
"args": {
|
| 2866 |
+
"should_epoch_stop": false,
|
| 2867 |
+
"should_evaluate": false,
|
| 2868 |
+
"should_log": false,
|
| 2869 |
+
"should_save": true,
|
| 2870 |
+
"should_training_stop": true
|
| 2871 |
+
},
|
| 2872 |
+
"attributes": {}
|
| 2873 |
+
}
|
| 2874 |
+
},
|
| 2875 |
+
"total_flos": 3.35021983285248e+19,
|
| 2876 |
+
"train_batch_size": 2,
|
| 2877 |
+
"trial_name": null,
|
| 2878 |
+
"trial_params": null
|
| 2879 |
+
}
|
checkpoint-200000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741392110574dc74949de4995bd5bf124fe829c8e6a7e3937ee5dcd15b77b966
|
| 3 |
+
size 5265
|
config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"dtype": "float32",
|
| 6 |
+
"id2label": {
|
| 7 |
+
"0": "ai",
|
| 8 |
+
"1": "hum"
|
| 9 |
+
},
|
| 10 |
+
"initializer_factor": 1.0,
|
| 11 |
+
"label2id": {
|
| 12 |
+
"ai": 0,
|
| 13 |
+
"hum": 1
|
| 14 |
+
},
|
| 15 |
+
"model_type": "siglip",
|
| 16 |
+
"problem_type": "single_label_classification",
|
| 17 |
+
"text_config": {
|
| 18 |
+
"attention_dropout": 0.0,
|
| 19 |
+
"dtype": "float32",
|
| 20 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"layer_norm_eps": 1e-06,
|
| 24 |
+
"max_position_embeddings": 64,
|
| 25 |
+
"model_type": "siglip_text_model",
|
| 26 |
+
"num_attention_heads": 12,
|
| 27 |
+
"num_hidden_layers": 12,
|
| 28 |
+
"projection_size": 768,
|
| 29 |
+
"vocab_size": 256000
|
| 30 |
+
},
|
| 31 |
+
"transformers_version": "5.0.0.dev0",
|
| 32 |
+
"use_cache": false,
|
| 33 |
+
"vision_config": {
|
| 34 |
+
"attention_dropout": 0.0,
|
| 35 |
+
"dtype": "float32",
|
| 36 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 37 |
+
"hidden_size": 768,
|
| 38 |
+
"image_size": 224,
|
| 39 |
+
"intermediate_size": 3072,
|
| 40 |
+
"layer_norm_eps": 1e-06,
|
| 41 |
+
"model_type": "siglip_vision_model",
|
| 42 |
+
"num_attention_heads": 12,
|
| 43 |
+
"num_channels": 3,
|
| 44 |
+
"num_hidden_layers": 12,
|
| 45 |
+
"patch_size": 16
|
| 46 |
+
}
|
| 47 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.98965,
|
| 4 |
+
"eval_loss": 0.10565917938947678,
|
| 5 |
+
"eval_runtime": 70.587,
|
| 6 |
+
"eval_samples_per_second": 283.338,
|
| 7 |
+
"eval_steps_per_second": 70.835
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:510e152b05df42e0fbd31fea37a63bf00b6be2749dec448b0c95810c67a7c94f
|
| 3 |
+
size 371567992
|
runs/Nov16_11-57-36_blacc/events.out.tfevents.1763283456.blacc
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cabd32a6cdc025db82207f185d53aa9c2e5c799139db636ea1c04835f2ef7b94
|
| 3 |
+
size 92470
|
runs/Nov16_11-57-36_blacc/events.out.tfevents.1763289551.blacc
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a570b1e386a94f2cc748f166d7f17c397ff03d1b55af36e16e7d2619f2d0ee0
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 3.35021983285248e+19,
|
| 4 |
+
"train_loss": 0.12970913526404707,
|
| 5 |
+
"train_runtime": 6024.3105,
|
| 6 |
+
"train_samples_per_second": 66.398,
|
| 7 |
+
"train_steps_per_second": 33.199
|
| 8 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741392110574dc74949de4995bd5bf124fe829c8e6a7e3937ee5dcd15b77b966
|
| 3 |
+
size 5265
|