CheckPoint16K
Browse filesNext version of TAMGA VLM
- config.json +54 -0
- generation_config.json +7 -0
- model.safetensors +3 -0
- trainer_state.json +1196 -0
config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BlipForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"image_text_hidden_size": 256,
|
| 6 |
+
"initializer_factor": 1.0,
|
| 7 |
+
"initializer_range": 0.02,
|
| 8 |
+
"label_smoothing": 0.0,
|
| 9 |
+
"logit_scale_init_value": 2.6592,
|
| 10 |
+
"model_type": "blip",
|
| 11 |
+
"projection_dim": 512,
|
| 12 |
+
"text_config": {
|
| 13 |
+
"_attn_implementation_autoset": true,
|
| 14 |
+
"attention_probs_dropout_prob": 0.0,
|
| 15 |
+
"encoder_hidden_size": 768,
|
| 16 |
+
"hidden_act": "gelu",
|
| 17 |
+
"hidden_dropout_prob": 0.0,
|
| 18 |
+
"hidden_size": 768,
|
| 19 |
+
"initializer_factor": 1.0,
|
| 20 |
+
"initializer_range": 0.02,
|
| 21 |
+
"intermediate_size": 3072,
|
| 22 |
+
"label_smoothing": 0.0,
|
| 23 |
+
"layer_norm_eps": 1e-12,
|
| 24 |
+
"max_position_embeddings": 512,
|
| 25 |
+
"model_type": "blip_text_model",
|
| 26 |
+
"num_attention_heads": 12,
|
| 27 |
+
"num_hidden_layers": 12,
|
| 28 |
+
"projection_dim": 768,
|
| 29 |
+
"torch_dtype": "float32",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 30524
|
| 32 |
+
},
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.51.3",
|
| 35 |
+
"vision_config": {
|
| 36 |
+
"_attn_implementation_autoset": true,
|
| 37 |
+
"attention_dropout": 0.0,
|
| 38 |
+
"dropout": 0.0,
|
| 39 |
+
"hidden_act": "gelu",
|
| 40 |
+
"hidden_size": 768,
|
| 41 |
+
"image_size": 384,
|
| 42 |
+
"initializer_factor": 1.0,
|
| 43 |
+
"initializer_range": 0.02,
|
| 44 |
+
"intermediate_size": 3072,
|
| 45 |
+
"layer_norm_eps": 1e-05,
|
| 46 |
+
"model_type": "blip_vision_model",
|
| 47 |
+
"num_attention_heads": 12,
|
| 48 |
+
"num_channels": 3,
|
| 49 |
+
"num_hidden_layers": 12,
|
| 50 |
+
"patch_size": 16,
|
| 51 |
+
"projection_dim": 512,
|
| 52 |
+
"torch_dtype": "float32"
|
| 53 |
+
}
|
| 54 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 30522,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.51.3"
|
| 7 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2875c081961facd9c6636391771b2f09abad3edd8a3484116bad2dc066f631a8
|
| 3 |
+
size 989717056
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1196 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 16664,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.04800768122899664,
|
| 14 |
+
"grad_norm": 3.5230019092559814,
|
| 15 |
+
"learning_rate": 4.995343254920787e-05,
|
| 16 |
+
"loss": 3.834,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.09601536245799328,
|
| 21 |
+
"grad_norm": 2.2122299671173096,
|
| 22 |
+
"learning_rate": 4.990542486797888e-05,
|
| 23 |
+
"loss": 1.006,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.14402304368698993,
|
| 28 |
+
"grad_norm": 1.6999797821044922,
|
| 29 |
+
"learning_rate": 4.985741718674988e-05,
|
| 30 |
+
"loss": 0.795,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.19203072491598655,
|
| 35 |
+
"grad_norm": 1.8523478507995605,
|
| 36 |
+
"learning_rate": 4.980940950552089e-05,
|
| 37 |
+
"loss": 0.7485,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.2400384061449832,
|
| 42 |
+
"grad_norm": 1.6911518573760986,
|
| 43 |
+
"learning_rate": 4.9761401824291885e-05,
|
| 44 |
+
"loss": 0.7051,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.28804608737397985,
|
| 49 |
+
"grad_norm": 1.78285551071167,
|
| 50 |
+
"learning_rate": 4.971339414306289e-05,
|
| 51 |
+
"loss": 0.6518,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.3360537686029765,
|
| 56 |
+
"grad_norm": 1.8025232553482056,
|
| 57 |
+
"learning_rate": 4.9665386461833895e-05,
|
| 58 |
+
"loss": 0.6429,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.3840614498319731,
|
| 63 |
+
"grad_norm": 1.6609948873519897,
|
| 64 |
+
"learning_rate": 4.96173787806049e-05,
|
| 65 |
+
"loss": 0.6148,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.4320691310609698,
|
| 70 |
+
"grad_norm": 1.7044763565063477,
|
| 71 |
+
"learning_rate": 4.9569371099375905e-05,
|
| 72 |
+
"loss": 0.6052,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.4800768122899664,
|
| 77 |
+
"grad_norm": 1.742863655090332,
|
| 78 |
+
"learning_rate": 4.9521363418146903e-05,
|
| 79 |
+
"loss": 0.599,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.528084493518963,
|
| 84 |
+
"grad_norm": 1.6556651592254639,
|
| 85 |
+
"learning_rate": 4.947335573691791e-05,
|
| 86 |
+
"loss": 0.5772,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.5760921747479597,
|
| 91 |
+
"grad_norm": 1.578718662261963,
|
| 92 |
+
"learning_rate": 4.9425348055688914e-05,
|
| 93 |
+
"loss": 0.5617,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.6240998559769563,
|
| 98 |
+
"grad_norm": 1.6328682899475098,
|
| 99 |
+
"learning_rate": 4.937734037445992e-05,
|
| 100 |
+
"loss": 0.5566,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.672107537205953,
|
| 105 |
+
"grad_norm": 1.775172233581543,
|
| 106 |
+
"learning_rate": 4.932981277004321e-05,
|
| 107 |
+
"loss": 0.5474,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.7201152184349496,
|
| 112 |
+
"grad_norm": 1.4435912370681763,
|
| 113 |
+
"learning_rate": 4.928180508881422e-05,
|
| 114 |
+
"loss": 0.5342,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.7681228996639462,
|
| 119 |
+
"grad_norm": 1.2456103563308716,
|
| 120 |
+
"learning_rate": 4.9233797407585215e-05,
|
| 121 |
+
"loss": 0.5237,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.8161305808929429,
|
| 126 |
+
"grad_norm": 1.5610588788986206,
|
| 127 |
+
"learning_rate": 4.918578972635622e-05,
|
| 128 |
+
"loss": 0.5335,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8641382621219396,
|
| 133 |
+
"grad_norm": 1.677467703819275,
|
| 134 |
+
"learning_rate": 4.9137782045127225e-05,
|
| 135 |
+
"loss": 0.5052,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.9121459433509361,
|
| 140 |
+
"grad_norm": 1.3189855813980103,
|
| 141 |
+
"learning_rate": 4.908977436389823e-05,
|
| 142 |
+
"loss": 0.4929,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.9601536245799328,
|
| 147 |
+
"grad_norm": 1.320538878440857,
|
| 148 |
+
"learning_rate": 4.904176668266923e-05,
|
| 149 |
+
"loss": 0.4973,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.0081613058089294,
|
| 154 |
+
"grad_norm": 1.2392332553863525,
|
| 155 |
+
"learning_rate": 4.8993759001440233e-05,
|
| 156 |
+
"loss": 0.4879,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.056168987037926,
|
| 161 |
+
"grad_norm": 1.2972139120101929,
|
| 162 |
+
"learning_rate": 4.894575132021124e-05,
|
| 163 |
+
"loss": 0.4339,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.1041766682669227,
|
| 168 |
+
"grad_norm": 1.763129711151123,
|
| 169 |
+
"learning_rate": 4.8897743638982243e-05,
|
| 170 |
+
"loss": 0.4215,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.1521843494959194,
|
| 175 |
+
"grad_norm": 1.8901598453521729,
|
| 176 |
+
"learning_rate": 4.884973595775325e-05,
|
| 177 |
+
"loss": 0.4386,
|
| 178 |
+
"step": 2400
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.2001920307249159,
|
| 182 |
+
"grad_norm": 1.1455730199813843,
|
| 183 |
+
"learning_rate": 4.880172827652425e-05,
|
| 184 |
+
"loss": 0.4414,
|
| 185 |
+
"step": 2500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.2481997119539125,
|
| 189 |
+
"grad_norm": 1.3231408596038818,
|
| 190 |
+
"learning_rate": 4.875372059529525e-05,
|
| 191 |
+
"loss": 0.4232,
|
| 192 |
+
"step": 2600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.2962073931829092,
|
| 196 |
+
"grad_norm": 1.544893503189087,
|
| 197 |
+
"learning_rate": 4.870571291406626e-05,
|
| 198 |
+
"loss": 0.4338,
|
| 199 |
+
"step": 2700
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.344215074411906,
|
| 203 |
+
"grad_norm": 1.4648122787475586,
|
| 204 |
+
"learning_rate": 4.8657705232837255e-05,
|
| 205 |
+
"loss": 0.4273,
|
| 206 |
+
"step": 2800
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.3922227556409026,
|
| 210 |
+
"grad_norm": 1.405002474784851,
|
| 211 |
+
"learning_rate": 4.860969755160826e-05,
|
| 212 |
+
"loss": 0.4337,
|
| 213 |
+
"step": 2900
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.4402304368698993,
|
| 217 |
+
"grad_norm": 1.2591880559921265,
|
| 218 |
+
"learning_rate": 4.8561689870379265e-05,
|
| 219 |
+
"loss": 0.4234,
|
| 220 |
+
"step": 3000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.488238118098896,
|
| 224 |
+
"grad_norm": 1.422620415687561,
|
| 225 |
+
"learning_rate": 4.8513682189150264e-05,
|
| 226 |
+
"loss": 0.4239,
|
| 227 |
+
"step": 3100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.5362457993278924,
|
| 231 |
+
"grad_norm": 1.585553526878357,
|
| 232 |
+
"learning_rate": 4.846567450792127e-05,
|
| 233 |
+
"loss": 0.4254,
|
| 234 |
+
"step": 3200
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.584253480556889,
|
| 238 |
+
"grad_norm": 1.1293108463287354,
|
| 239 |
+
"learning_rate": 4.8417666826692274e-05,
|
| 240 |
+
"loss": 0.422,
|
| 241 |
+
"step": 3300
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.6322611617858858,
|
| 245 |
+
"grad_norm": 1.265154242515564,
|
| 246 |
+
"learning_rate": 4.836965914546328e-05,
|
| 247 |
+
"loss": 0.4194,
|
| 248 |
+
"step": 3400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.6802688430148822,
|
| 252 |
+
"grad_norm": 1.381218433380127,
|
| 253 |
+
"learning_rate": 4.832165146423428e-05,
|
| 254 |
+
"loss": 0.4321,
|
| 255 |
+
"step": 3500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.7282765242438791,
|
| 259 |
+
"grad_norm": 1.3752981424331665,
|
| 260 |
+
"learning_rate": 4.827364378300528e-05,
|
| 261 |
+
"loss": 0.4151,
|
| 262 |
+
"step": 3600
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.7762842054728756,
|
| 266 |
+
"grad_norm": 1.1908366680145264,
|
| 267 |
+
"learning_rate": 4.822563610177629e-05,
|
| 268 |
+
"loss": 0.4203,
|
| 269 |
+
"step": 3700
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.8242918867018723,
|
| 273 |
+
"grad_norm": 1.2795414924621582,
|
| 274 |
+
"learning_rate": 4.8177628420547285e-05,
|
| 275 |
+
"loss": 0.414,
|
| 276 |
+
"step": 3800
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.872299567930869,
|
| 280 |
+
"grad_norm": 1.2664345502853394,
|
| 281 |
+
"learning_rate": 4.812962073931829e-05,
|
| 282 |
+
"loss": 0.4102,
|
| 283 |
+
"step": 3900
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.9203072491598656,
|
| 287 |
+
"grad_norm": 1.1444271802902222,
|
| 288 |
+
"learning_rate": 4.8081613058089295e-05,
|
| 289 |
+
"loss": 0.4066,
|
| 290 |
+
"step": 4000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.9683149303888623,
|
| 294 |
+
"grad_norm": 1.233847737312317,
|
| 295 |
+
"learning_rate": 4.80336053768603e-05,
|
| 296 |
+
"loss": 0.4235,
|
| 297 |
+
"step": 4100
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.0163226116178588,
|
| 301 |
+
"grad_norm": 1.1085718870162964,
|
| 302 |
+
"learning_rate": 4.79855976956313e-05,
|
| 303 |
+
"loss": 0.3852,
|
| 304 |
+
"step": 4200
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.0643302928468557,
|
| 308 |
+
"grad_norm": 1.2250890731811523,
|
| 309 |
+
"learning_rate": 4.7937590014402304e-05,
|
| 310 |
+
"loss": 0.334,
|
| 311 |
+
"step": 4300
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.112337974075852,
|
| 315 |
+
"grad_norm": 1.3685556650161743,
|
| 316 |
+
"learning_rate": 4.788958233317331e-05,
|
| 317 |
+
"loss": 0.3273,
|
| 318 |
+
"step": 4400
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.160345655304849,
|
| 322 |
+
"grad_norm": 1.2295827865600586,
|
| 323 |
+
"learning_rate": 4.7841574651944314e-05,
|
| 324 |
+
"loss": 0.3332,
|
| 325 |
+
"step": 4500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.2083533365338455,
|
| 329 |
+
"grad_norm": 1.3651865720748901,
|
| 330 |
+
"learning_rate": 4.779356697071531e-05,
|
| 331 |
+
"loss": 0.3355,
|
| 332 |
+
"step": 4600
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.256361017762842,
|
| 336 |
+
"grad_norm": 1.414611577987671,
|
| 337 |
+
"learning_rate": 4.774555928948632e-05,
|
| 338 |
+
"loss": 0.3347,
|
| 339 |
+
"step": 4700
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.304368698991839,
|
| 343 |
+
"grad_norm": 1.1170624494552612,
|
| 344 |
+
"learning_rate": 4.769755160825732e-05,
|
| 345 |
+
"loss": 0.3363,
|
| 346 |
+
"step": 4800
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.3523763802208353,
|
| 350 |
+
"grad_norm": 1.4784187078475952,
|
| 351 |
+
"learning_rate": 4.764954392702833e-05,
|
| 352 |
+
"loss": 0.3497,
|
| 353 |
+
"step": 4900
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.4003840614498317,
|
| 357 |
+
"grad_norm": 1.3222328424453735,
|
| 358 |
+
"learning_rate": 4.7601536245799325e-05,
|
| 359 |
+
"loss": 0.3319,
|
| 360 |
+
"step": 5000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.4483917426788286,
|
| 364 |
+
"grad_norm": 1.4248158931732178,
|
| 365 |
+
"learning_rate": 4.755352856457033e-05,
|
| 366 |
+
"loss": 0.3456,
|
| 367 |
+
"step": 5100
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.496399423907825,
|
| 371 |
+
"grad_norm": 1.3521333932876587,
|
| 372 |
+
"learning_rate": 4.7505520883341336e-05,
|
| 373 |
+
"loss": 0.346,
|
| 374 |
+
"step": 5200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.544407105136822,
|
| 378 |
+
"grad_norm": 1.3070430755615234,
|
| 379 |
+
"learning_rate": 4.745751320211234e-05,
|
| 380 |
+
"loss": 0.338,
|
| 381 |
+
"step": 5300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.5924147863658185,
|
| 385 |
+
"grad_norm": 1.2667192220687866,
|
| 386 |
+
"learning_rate": 4.7409505520883346e-05,
|
| 387 |
+
"loss": 0.3508,
|
| 388 |
+
"step": 5400
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.6404224675948154,
|
| 392 |
+
"grad_norm": 1.404773473739624,
|
| 393 |
+
"learning_rate": 4.7361497839654344e-05,
|
| 394 |
+
"loss": 0.3376,
|
| 395 |
+
"step": 5500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 2.688430148823812,
|
| 399 |
+
"grad_norm": 1.233578085899353,
|
| 400 |
+
"learning_rate": 4.731349015842535e-05,
|
| 401 |
+
"loss": 0.3432,
|
| 402 |
+
"step": 5600
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.7364378300528083,
|
| 406 |
+
"grad_norm": 1.2811566591262817,
|
| 407 |
+
"learning_rate": 4.7265482477196354e-05,
|
| 408 |
+
"loss": 0.3498,
|
| 409 |
+
"step": 5700
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.784445511281805,
|
| 413 |
+
"grad_norm": 1.1714744567871094,
|
| 414 |
+
"learning_rate": 4.721747479596736e-05,
|
| 415 |
+
"loss": 0.3451,
|
| 416 |
+
"step": 5800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.8324531925108016,
|
| 420 |
+
"grad_norm": 1.1416313648223877,
|
| 421 |
+
"learning_rate": 4.716946711473836e-05,
|
| 422 |
+
"loss": 0.3427,
|
| 423 |
+
"step": 5900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.8804608737397985,
|
| 427 |
+
"grad_norm": 1.3504301309585571,
|
| 428 |
+
"learning_rate": 4.712145943350936e-05,
|
| 429 |
+
"loss": 0.3477,
|
| 430 |
+
"step": 6000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.928468554968795,
|
| 434 |
+
"grad_norm": 1.4579728841781616,
|
| 435 |
+
"learning_rate": 4.707345175228037e-05,
|
| 436 |
+
"loss": 0.3473,
|
| 437 |
+
"step": 6100
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.976476236197792,
|
| 441 |
+
"grad_norm": 1.1714892387390137,
|
| 442 |
+
"learning_rate": 4.702544407105137e-05,
|
| 443 |
+
"loss": 0.341,
|
| 444 |
+
"step": 6200
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.0244839174267883,
|
| 448 |
+
"grad_norm": 1.200076699256897,
|
| 449 |
+
"learning_rate": 4.697743638982237e-05,
|
| 450 |
+
"loss": 0.2989,
|
| 451 |
+
"step": 6300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.072491598655785,
|
| 455 |
+
"grad_norm": 1.3686045408248901,
|
| 456 |
+
"learning_rate": 4.6929428708593376e-05,
|
| 457 |
+
"loss": 0.2606,
|
| 458 |
+
"step": 6400
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.1204992798847817,
|
| 462 |
+
"grad_norm": 1.2546569108963013,
|
| 463 |
+
"learning_rate": 4.688142102736438e-05,
|
| 464 |
+
"loss": 0.2635,
|
| 465 |
+
"step": 6500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.168506961113778,
|
| 469 |
+
"grad_norm": 1.3534624576568604,
|
| 470 |
+
"learning_rate": 4.6833413346135386e-05,
|
| 471 |
+
"loss": 0.2606,
|
| 472 |
+
"step": 6600
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 3.2165146423427746,
|
| 476 |
+
"grad_norm": 1.4877707958221436,
|
| 477 |
+
"learning_rate": 4.678540566490639e-05,
|
| 478 |
+
"loss": 0.2685,
|
| 479 |
+
"step": 6700
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.2645223235717715,
|
| 483 |
+
"grad_norm": 1.4656774997711182,
|
| 484 |
+
"learning_rate": 4.673739798367739e-05,
|
| 485 |
+
"loss": 0.272,
|
| 486 |
+
"step": 6800
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.312530004800768,
|
| 490 |
+
"grad_norm": 1.24815034866333,
|
| 491 |
+
"learning_rate": 4.6689390302448394e-05,
|
| 492 |
+
"loss": 0.2703,
|
| 493 |
+
"step": 6900
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 3.360537686029765,
|
| 497 |
+
"grad_norm": 1.1822447776794434,
|
| 498 |
+
"learning_rate": 4.66413826212194e-05,
|
| 499 |
+
"loss": 0.2702,
|
| 500 |
+
"step": 7000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.4085453672587613,
|
| 504 |
+
"grad_norm": 1.4170218706130981,
|
| 505 |
+
"learning_rate": 4.6593374939990404e-05,
|
| 506 |
+
"loss": 0.2716,
|
| 507 |
+
"step": 7100
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.4565530484877582,
|
| 511 |
+
"grad_norm": 1.360338568687439,
|
| 512 |
+
"learning_rate": 4.65453672587614e-05,
|
| 513 |
+
"loss": 0.2717,
|
| 514 |
+
"step": 7200
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 3.5045607297167547,
|
| 518 |
+
"grad_norm": 1.4744322299957275,
|
| 519 |
+
"learning_rate": 4.649735957753241e-05,
|
| 520 |
+
"loss": 0.2795,
|
| 521 |
+
"step": 7300
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.552568410945751,
|
| 525 |
+
"grad_norm": 1.5080569982528687,
|
| 526 |
+
"learning_rate": 4.644935189630341e-05,
|
| 527 |
+
"loss": 0.2753,
|
| 528 |
+
"step": 7400
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 3.600576092174748,
|
| 532 |
+
"grad_norm": 1.5707645416259766,
|
| 533 |
+
"learning_rate": 4.640134421507442e-05,
|
| 534 |
+
"loss": 0.278,
|
| 535 |
+
"step": 7500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 3.6485837734037445,
|
| 539 |
+
"grad_norm": 1.4137930870056152,
|
| 540 |
+
"learning_rate": 4.6353336533845416e-05,
|
| 541 |
+
"loss": 0.2756,
|
| 542 |
+
"step": 7600
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.696591454632741,
|
| 546 |
+
"grad_norm": 1.4156709909439087,
|
| 547 |
+
"learning_rate": 4.630532885261642e-05,
|
| 548 |
+
"loss": 0.2841,
|
| 549 |
+
"step": 7700
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.744599135861738,
|
| 553 |
+
"grad_norm": 1.4720118045806885,
|
| 554 |
+
"learning_rate": 4.6257321171387426e-05,
|
| 555 |
+
"loss": 0.2767,
|
| 556 |
+
"step": 7800
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.7926068170907348,
|
| 560 |
+
"grad_norm": 1.2695891857147217,
|
| 561 |
+
"learning_rate": 4.620931349015843e-05,
|
| 562 |
+
"loss": 0.2762,
|
| 563 |
+
"step": 7900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.8406144983197312,
|
| 567 |
+
"grad_norm": 1.439794659614563,
|
| 568 |
+
"learning_rate": 4.6161305808929436e-05,
|
| 569 |
+
"loss": 0.2799,
|
| 570 |
+
"step": 8000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.8886221795487277,
|
| 574 |
+
"grad_norm": 1.3036625385284424,
|
| 575 |
+
"learning_rate": 4.611377820451272e-05,
|
| 576 |
+
"loss": 0.2728,
|
| 577 |
+
"step": 8100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 3.9366298607777246,
|
| 581 |
+
"grad_norm": 1.3942623138427734,
|
| 582 |
+
"learning_rate": 4.606577052328373e-05,
|
| 583 |
+
"loss": 0.2779,
|
| 584 |
+
"step": 8200
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.984637542006721,
|
| 588 |
+
"grad_norm": 1.3402965068817139,
|
| 589 |
+
"learning_rate": 4.6017762842054726e-05,
|
| 590 |
+
"loss": 0.2759,
|
| 591 |
+
"step": 8300
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.0326452232357175,
|
| 595 |
+
"grad_norm": 1.2347018718719482,
|
| 596 |
+
"learning_rate": 4.596975516082573e-05,
|
| 597 |
+
"loss": 0.2239,
|
| 598 |
+
"step": 8400
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.080652904464714,
|
| 602 |
+
"grad_norm": 1.2590075731277466,
|
| 603 |
+
"learning_rate": 4.5922227556409024e-05,
|
| 604 |
+
"loss": 0.1964,
|
| 605 |
+
"step": 8500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.128660585693711,
|
| 609 |
+
"grad_norm": 1.2734559774398804,
|
| 610 |
+
"learning_rate": 4.587421987518003e-05,
|
| 611 |
+
"loss": 0.2033,
|
| 612 |
+
"step": 8600
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 4.176668266922707,
|
| 616 |
+
"grad_norm": 1.3685240745544434,
|
| 617 |
+
"learning_rate": 4.5826212193951034e-05,
|
| 618 |
+
"loss": 0.2039,
|
| 619 |
+
"step": 8700
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 4.224675948151704,
|
| 623 |
+
"grad_norm": 1.2431707382202148,
|
| 624 |
+
"learning_rate": 4.577820451272204e-05,
|
| 625 |
+
"loss": 0.2031,
|
| 626 |
+
"step": 8800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.272683629380701,
|
| 630 |
+
"grad_norm": 1.3057847023010254,
|
| 631 |
+
"learning_rate": 4.573019683149304e-05,
|
| 632 |
+
"loss": 0.1993,
|
| 633 |
+
"step": 8900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.320691310609698,
|
| 637 |
+
"grad_norm": 1.3426510095596313,
|
| 638 |
+
"learning_rate": 4.568218915026404e-05,
|
| 639 |
+
"loss": 0.2103,
|
| 640 |
+
"step": 9000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 4.368698991838694,
|
| 644 |
+
"grad_norm": 1.4993879795074463,
|
| 645 |
+
"learning_rate": 4.563418146903505e-05,
|
| 646 |
+
"loss": 0.2038,
|
| 647 |
+
"step": 9100
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 4.416706673067691,
|
| 651 |
+
"grad_norm": 1.1477452516555786,
|
| 652 |
+
"learning_rate": 4.558617378780605e-05,
|
| 653 |
+
"loss": 0.2087,
|
| 654 |
+
"step": 9200
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 4.464714354296688,
|
| 658 |
+
"grad_norm": 1.2744262218475342,
|
| 659 |
+
"learning_rate": 4.553816610657706e-05,
|
| 660 |
+
"loss": 0.2128,
|
| 661 |
+
"step": 9300
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 4.512722035525684,
|
| 665 |
+
"grad_norm": 1.2372483015060425,
|
| 666 |
+
"learning_rate": 4.5490158425348056e-05,
|
| 667 |
+
"loss": 0.2141,
|
| 668 |
+
"step": 9400
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.560729716754681,
|
| 672 |
+
"grad_norm": 1.5185002088546753,
|
| 673 |
+
"learning_rate": 4.544215074411906e-05,
|
| 674 |
+
"loss": 0.2163,
|
| 675 |
+
"step": 9500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 4.608737397983678,
|
| 679 |
+
"grad_norm": 1.2318006753921509,
|
| 680 |
+
"learning_rate": 4.5394143062890066e-05,
|
| 681 |
+
"loss": 0.2142,
|
| 682 |
+
"step": 9600
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 4.656745079212674,
|
| 686 |
+
"grad_norm": 1.0812604427337646,
|
| 687 |
+
"learning_rate": 4.534613538166107e-05,
|
| 688 |
+
"loss": 0.2179,
|
| 689 |
+
"step": 9700
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 4.704752760441671,
|
| 693 |
+
"grad_norm": 1.460476279258728,
|
| 694 |
+
"learning_rate": 4.529812770043207e-05,
|
| 695 |
+
"loss": 0.2203,
|
| 696 |
+
"step": 9800
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 4.7527604416706675,
|
| 700 |
+
"grad_norm": 1.4925509691238403,
|
| 701 |
+
"learning_rate": 4.5250120019203074e-05,
|
| 702 |
+
"loss": 0.2167,
|
| 703 |
+
"step": 9900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 4.8007681228996635,
|
| 707 |
+
"grad_norm": 1.4219658374786377,
|
| 708 |
+
"learning_rate": 4.520211233797408e-05,
|
| 709 |
+
"loss": 0.2163,
|
| 710 |
+
"step": 10000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.84877580412866,
|
| 714 |
+
"grad_norm": 1.0717830657958984,
|
| 715 |
+
"learning_rate": 4.5154104656745084e-05,
|
| 716 |
+
"loss": 0.2138,
|
| 717 |
+
"step": 10100
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 4.896783485357657,
|
| 721 |
+
"grad_norm": 1.0514252185821533,
|
| 722 |
+
"learning_rate": 4.510609697551608e-05,
|
| 723 |
+
"loss": 0.2187,
|
| 724 |
+
"step": 10200
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 4.944791166586654,
|
| 728 |
+
"grad_norm": 1.1718051433563232,
|
| 729 |
+
"learning_rate": 4.505808929428709e-05,
|
| 730 |
+
"loss": 0.2179,
|
| 731 |
+
"step": 10300
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 4.99279884781565,
|
| 735 |
+
"grad_norm": 1.3955721855163574,
|
| 736 |
+
"learning_rate": 4.501008161305809e-05,
|
| 737 |
+
"loss": 0.2222,
|
| 738 |
+
"step": 10400
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 5.040806529044647,
|
| 742 |
+
"grad_norm": 1.0917434692382812,
|
| 743 |
+
"learning_rate": 4.49620739318291e-05,
|
| 744 |
+
"loss": 0.1556,
|
| 745 |
+
"step": 10500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 5.088814210273644,
|
| 749 |
+
"grad_norm": 1.2517110109329224,
|
| 750 |
+
"learning_rate": 4.49140662506001e-05,
|
| 751 |
+
"loss": 0.1442,
|
| 752 |
+
"step": 10600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 5.13682189150264,
|
| 756 |
+
"grad_norm": 1.4139866828918457,
|
| 757 |
+
"learning_rate": 4.48660585693711e-05,
|
| 758 |
+
"loss": 0.1456,
|
| 759 |
+
"step": 10700
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 5.184829572731637,
|
| 763 |
+
"grad_norm": 1.4004515409469604,
|
| 764 |
+
"learning_rate": 4.4818050888142106e-05,
|
| 765 |
+
"loss": 0.1475,
|
| 766 |
+
"step": 10800
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 5.232837253960634,
|
| 770 |
+
"grad_norm": 1.3576511144638062,
|
| 771 |
+
"learning_rate": 4.477004320691311e-05,
|
| 772 |
+
"loss": 0.1475,
|
| 773 |
+
"step": 10900
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 5.280844935189631,
|
| 777 |
+
"grad_norm": 1.2428905963897705,
|
| 778 |
+
"learning_rate": 4.4722035525684116e-05,
|
| 779 |
+
"loss": 0.1505,
|
| 780 |
+
"step": 11000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 5.328852616418627,
|
| 784 |
+
"grad_norm": 1.577462077140808,
|
| 785 |
+
"learning_rate": 4.4674027844455114e-05,
|
| 786 |
+
"loss": 0.1533,
|
| 787 |
+
"step": 11100
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 5.376860297647624,
|
| 791 |
+
"grad_norm": 1.5575320720672607,
|
| 792 |
+
"learning_rate": 4.462602016322612e-05,
|
| 793 |
+
"loss": 0.1545,
|
| 794 |
+
"step": 11200
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 5.4248679788766205,
|
| 798 |
+
"grad_norm": 1.6392663717269897,
|
| 799 |
+
"learning_rate": 4.4578012481997124e-05,
|
| 800 |
+
"loss": 0.1535,
|
| 801 |
+
"step": 11300
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 5.4728756601056165,
|
| 805 |
+
"grad_norm": 1.5311086177825928,
|
| 806 |
+
"learning_rate": 4.453000480076813e-05,
|
| 807 |
+
"loss": 0.1579,
|
| 808 |
+
"step": 11400
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 5.5208833413346134,
|
| 812 |
+
"grad_norm": 1.2279739379882812,
|
| 813 |
+
"learning_rate": 4.448199711953913e-05,
|
| 814 |
+
"loss": 0.1616,
|
| 815 |
+
"step": 11500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 5.56889102256361,
|
| 819 |
+
"grad_norm": 1.48116934299469,
|
| 820 |
+
"learning_rate": 4.443398943831013e-05,
|
| 821 |
+
"loss": 0.1603,
|
| 822 |
+
"step": 11600
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 5.616898703792607,
|
| 826 |
+
"grad_norm": 1.2573692798614502,
|
| 827 |
+
"learning_rate": 4.438598175708114e-05,
|
| 828 |
+
"loss": 0.1615,
|
| 829 |
+
"step": 11700
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 5.664906385021603,
|
| 833 |
+
"grad_norm": 1.3175160884857178,
|
| 834 |
+
"learning_rate": 4.433797407585214e-05,
|
| 835 |
+
"loss": 0.1618,
|
| 836 |
+
"step": 11800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 5.7129140662506,
|
| 840 |
+
"grad_norm": 1.3634432554244995,
|
| 841 |
+
"learning_rate": 4.428996639462314e-05,
|
| 842 |
+
"loss": 0.1634,
|
| 843 |
+
"step": 11900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 5.760921747479597,
|
| 847 |
+
"grad_norm": 1.4681994915008545,
|
| 848 |
+
"learning_rate": 4.4241958713394146e-05,
|
| 849 |
+
"loss": 0.1644,
|
| 850 |
+
"step": 12000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 5.808929428708593,
|
| 854 |
+
"grad_norm": 2.1259870529174805,
|
| 855 |
+
"learning_rate": 4.419395103216515e-05,
|
| 856 |
+
"loss": 0.1637,
|
| 857 |
+
"step": 12100
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 5.85693710993759,
|
| 861 |
+
"grad_norm": 1.3507694005966187,
|
| 862 |
+
"learning_rate": 4.4145943350936156e-05,
|
| 863 |
+
"loss": 0.1676,
|
| 864 |
+
"step": 12200
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 5.904944791166587,
|
| 868 |
+
"grad_norm": 1.3994966745376587,
|
| 869 |
+
"learning_rate": 4.4097935669707155e-05,
|
| 870 |
+
"loss": 0.1682,
|
| 871 |
+
"step": 12300
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 5.952952472395584,
|
| 875 |
+
"grad_norm": 1.37490713596344,
|
| 876 |
+
"learning_rate": 4.404992798847816e-05,
|
| 877 |
+
"loss": 0.1632,
|
| 878 |
+
"step": 12400
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 6.00096015362458,
|
| 882 |
+
"grad_norm": 0.9519465565681458,
|
| 883 |
+
"learning_rate": 4.4001920307249165e-05,
|
| 884 |
+
"loss": 0.1647,
|
| 885 |
+
"step": 12500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 6.048967834853577,
|
| 889 |
+
"grad_norm": 1.1763761043548584,
|
| 890 |
+
"learning_rate": 4.395391262602016e-05,
|
| 891 |
+
"loss": 0.1012,
|
| 892 |
+
"step": 12600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 6.096975516082574,
|
| 896 |
+
"grad_norm": 1.311137318611145,
|
| 897 |
+
"learning_rate": 4.390590494479117e-05,
|
| 898 |
+
"loss": 0.1023,
|
| 899 |
+
"step": 12700
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 6.14498319731157,
|
| 903 |
+
"grad_norm": 0.9711358547210693,
|
| 904 |
+
"learning_rate": 4.385789726356217e-05,
|
| 905 |
+
"loss": 0.1025,
|
| 906 |
+
"step": 12800
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 6.1929908785405665,
|
| 910 |
+
"grad_norm": 1.7737120389938354,
|
| 911 |
+
"learning_rate": 4.380988958233317e-05,
|
| 912 |
+
"loss": 0.107,
|
| 913 |
+
"step": 12900
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 6.240998559769563,
|
| 917 |
+
"grad_norm": 1.3438420295715332,
|
| 918 |
+
"learning_rate": 4.3761881901104176e-05,
|
| 919 |
+
"loss": 0.1116,
|
| 920 |
+
"step": 13000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 6.289006240998559,
|
| 924 |
+
"grad_norm": 1.2701375484466553,
|
| 925 |
+
"learning_rate": 4.371387421987518e-05,
|
| 926 |
+
"loss": 0.1102,
|
| 927 |
+
"step": 13100
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 6.337013922227556,
|
| 931 |
+
"grad_norm": 1.3952118158340454,
|
| 932 |
+
"learning_rate": 4.366586653864618e-05,
|
| 933 |
+
"loss": 0.1114,
|
| 934 |
+
"step": 13200
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 6.385021603456553,
|
| 938 |
+
"grad_norm": 1.2892017364501953,
|
| 939 |
+
"learning_rate": 4.3617858857417185e-05,
|
| 940 |
+
"loss": 0.111,
|
| 941 |
+
"step": 13300
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 6.433029284685549,
|
| 945 |
+
"grad_norm": 1.2740648984909058,
|
| 946 |
+
"learning_rate": 4.356985117618819e-05,
|
| 947 |
+
"loss": 0.1125,
|
| 948 |
+
"step": 13400
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 6.481036965914546,
|
| 952 |
+
"grad_norm": 1.1152820587158203,
|
| 953 |
+
"learning_rate": 4.3521843494959195e-05,
|
| 954 |
+
"loss": 0.1145,
|
| 955 |
+
"step": 13500
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 6.529044647143543,
|
| 959 |
+
"grad_norm": 1.3018417358398438,
|
| 960 |
+
"learning_rate": 4.34738358137302e-05,
|
| 961 |
+
"loss": 0.1161,
|
| 962 |
+
"step": 13600
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 6.57705232837254,
|
| 966 |
+
"grad_norm": 1.3537256717681885,
|
| 967 |
+
"learning_rate": 4.34258281325012e-05,
|
| 968 |
+
"loss": 0.1127,
|
| 969 |
+
"step": 13700
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 6.625060009601536,
|
| 973 |
+
"grad_norm": 1.566361665725708,
|
| 974 |
+
"learning_rate": 4.33778204512722e-05,
|
| 975 |
+
"loss": 0.1168,
|
| 976 |
+
"step": 13800
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 6.673067690830533,
|
| 980 |
+
"grad_norm": 1.4471561908721924,
|
| 981 |
+
"learning_rate": 4.332981277004321e-05,
|
| 982 |
+
"loss": 0.1185,
|
| 983 |
+
"step": 13900
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 6.72107537205953,
|
| 987 |
+
"grad_norm": 1.2443643808364868,
|
| 988 |
+
"learning_rate": 4.328180508881421e-05,
|
| 989 |
+
"loss": 0.1211,
|
| 990 |
+
"step": 14000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 6.769083053288526,
|
| 994 |
+
"grad_norm": 1.568220615386963,
|
| 995 |
+
"learning_rate": 4.323379740758521e-05,
|
| 996 |
+
"loss": 0.1181,
|
| 997 |
+
"step": 14100
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 6.817090734517523,
|
| 1001 |
+
"grad_norm": 1.3633244037628174,
|
| 1002 |
+
"learning_rate": 4.3185789726356216e-05,
|
| 1003 |
+
"loss": 0.1221,
|
| 1004 |
+
"step": 14200
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 6.86509841574652,
|
| 1008 |
+
"grad_norm": 1.2666867971420288,
|
| 1009 |
+
"learning_rate": 4.313778204512722e-05,
|
| 1010 |
+
"loss": 0.1223,
|
| 1011 |
+
"step": 14300
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 6.9131060969755165,
|
| 1015 |
+
"grad_norm": 1.2681658267974854,
|
| 1016 |
+
"learning_rate": 4.3089774363898227e-05,
|
| 1017 |
+
"loss": 0.1231,
|
| 1018 |
+
"step": 14400
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 6.9611137782045125,
|
| 1022 |
+
"grad_norm": 1.0958737134933472,
|
| 1023 |
+
"learning_rate": 4.3041766682669225e-05,
|
| 1024 |
+
"loss": 0.123,
|
| 1025 |
+
"step": 14500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 7.009121459433509,
|
| 1029 |
+
"grad_norm": 0.9932379126548767,
|
| 1030 |
+
"learning_rate": 4.299375900144023e-05,
|
| 1031 |
+
"loss": 0.1141,
|
| 1032 |
+
"step": 14600
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.057129140662506,
|
| 1036 |
+
"grad_norm": 1.1502649784088135,
|
| 1037 |
+
"learning_rate": 4.2945751320211235e-05,
|
| 1038 |
+
"loss": 0.0742,
|
| 1039 |
+
"step": 14700
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 7.105136821891502,
|
| 1043 |
+
"grad_norm": 1.1436489820480347,
|
| 1044 |
+
"learning_rate": 4.289774363898224e-05,
|
| 1045 |
+
"loss": 0.0751,
|
| 1046 |
+
"step": 14800
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 7.153144503120499,
|
| 1050 |
+
"grad_norm": 1.4604742527008057,
|
| 1051 |
+
"learning_rate": 4.2849735957753245e-05,
|
| 1052 |
+
"loss": 0.0766,
|
| 1053 |
+
"step": 14900
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 7.201152184349496,
|
| 1057 |
+
"grad_norm": 1.307935118675232,
|
| 1058 |
+
"learning_rate": 4.280172827652424e-05,
|
| 1059 |
+
"loss": 0.0784,
|
| 1060 |
+
"step": 15000
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 7.249159865578493,
|
| 1064 |
+
"grad_norm": 1.0527863502502441,
|
| 1065 |
+
"learning_rate": 4.275372059529525e-05,
|
| 1066 |
+
"loss": 0.0796,
|
| 1067 |
+
"step": 15100
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 7.297167546807489,
|
| 1071 |
+
"grad_norm": 1.2827943563461304,
|
| 1072 |
+
"learning_rate": 4.270571291406625e-05,
|
| 1073 |
+
"loss": 0.0783,
|
| 1074 |
+
"step": 15200
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 7.345175228036486,
|
| 1078 |
+
"grad_norm": 1.3460379838943481,
|
| 1079 |
+
"learning_rate": 4.265770523283726e-05,
|
| 1080 |
+
"loss": 0.0819,
|
| 1081 |
+
"step": 15300
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 7.393182909265483,
|
| 1085 |
+
"grad_norm": 1.2280845642089844,
|
| 1086 |
+
"learning_rate": 4.260969755160826e-05,
|
| 1087 |
+
"loss": 0.0843,
|
| 1088 |
+
"step": 15400
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 7.441190590494479,
|
| 1092 |
+
"grad_norm": 1.3223485946655273,
|
| 1093 |
+
"learning_rate": 4.256168987037926e-05,
|
| 1094 |
+
"loss": 0.0817,
|
| 1095 |
+
"step": 15500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 7.489198271723476,
|
| 1099 |
+
"grad_norm": 1.296045184135437,
|
| 1100 |
+
"learning_rate": 4.251368218915027e-05,
|
| 1101 |
+
"loss": 0.0832,
|
| 1102 |
+
"step": 15600
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 7.537205952952473,
|
| 1106 |
+
"grad_norm": 1.3680839538574219,
|
| 1107 |
+
"learning_rate": 4.246567450792127e-05,
|
| 1108 |
+
"loss": 0.0861,
|
| 1109 |
+
"step": 15700
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 7.585213634181469,
|
| 1113 |
+
"grad_norm": 1.1178674697875977,
|
| 1114 |
+
"learning_rate": 4.241766682669227e-05,
|
| 1115 |
+
"loss": 0.0838,
|
| 1116 |
+
"step": 15800
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 7.633221315410466,
|
| 1120 |
+
"grad_norm": 1.034132719039917,
|
| 1121 |
+
"learning_rate": 4.2369659145463275e-05,
|
| 1122 |
+
"loss": 0.0884,
|
| 1123 |
+
"step": 15900
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 7.6812289966394625,
|
| 1127 |
+
"grad_norm": 1.2711660861968994,
|
| 1128 |
+
"learning_rate": 4.232165146423428e-05,
|
| 1129 |
+
"loss": 0.0881,
|
| 1130 |
+
"step": 16000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 7.7292366778684585,
|
| 1134 |
+
"grad_norm": 1.1275447607040405,
|
| 1135 |
+
"learning_rate": 4.2273643783005285e-05,
|
| 1136 |
+
"loss": 0.0891,
|
| 1137 |
+
"step": 16100
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 7.777244359097455,
|
| 1141 |
+
"grad_norm": 1.1659106016159058,
|
| 1142 |
+
"learning_rate": 4.2225636101776283e-05,
|
| 1143 |
+
"loss": 0.0905,
|
| 1144 |
+
"step": 16200
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 7.825252040326452,
|
| 1148 |
+
"grad_norm": 1.3686105012893677,
|
| 1149 |
+
"learning_rate": 4.217762842054729e-05,
|
| 1150 |
+
"loss": 0.0911,
|
| 1151 |
+
"step": 16300
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 7.873259721555449,
|
| 1155 |
+
"grad_norm": 1.3593393564224243,
|
| 1156 |
+
"learning_rate": 4.2129620739318294e-05,
|
| 1157 |
+
"loss": 0.0917,
|
| 1158 |
+
"step": 16400
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 7.921267402784445,
|
| 1162 |
+
"grad_norm": 1.4164599180221558,
|
| 1163 |
+
"learning_rate": 4.20816130580893e-05,
|
| 1164 |
+
"loss": 0.092,
|
| 1165 |
+
"step": 16500
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 7.969275084013442,
|
| 1169 |
+
"grad_norm": 1.3200361728668213,
|
| 1170 |
+
"learning_rate": 4.2033605376860304e-05,
|
| 1171 |
+
"loss": 0.0926,
|
| 1172 |
+
"step": 16600
|
| 1173 |
+
}
|
| 1174 |
+
],
|
| 1175 |
+
"logging_steps": 100,
|
| 1176 |
+
"max_steps": 104150,
|
| 1177 |
+
"num_input_tokens_seen": 0,
|
| 1178 |
+
"num_train_epochs": 50,
|
| 1179 |
+
"save_steps": 10000,
|
| 1180 |
+
"stateful_callbacks": {
|
| 1181 |
+
"TrainerControl": {
|
| 1182 |
+
"args": {
|
| 1183 |
+
"should_epoch_stop": false,
|
| 1184 |
+
"should_evaluate": false,
|
| 1185 |
+
"should_log": false,
|
| 1186 |
+
"should_save": true,
|
| 1187 |
+
"should_training_stop": false
|
| 1188 |
+
},
|
| 1189 |
+
"attributes": {}
|
| 1190 |
+
}
|
| 1191 |
+
},
|
| 1192 |
+
"total_flos": 1.582019982939736e+20,
|
| 1193 |
+
"train_batch_size": 16,
|
| 1194 |
+
"trial_name": null,
|
| 1195 |
+
"trial_params": null
|
| 1196 |
+
}
|