Initial model upload
Browse files- config.json +60 -0
- generation_config.json +7 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2246 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"T5ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"classifier_dropout": 0.0,
|
| 6 |
+
"d_ff": 3072,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 768,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "relu",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "relu",
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"is_encoder_decoder": true,
|
| 16 |
+
"is_gated_act": false,
|
| 17 |
+
"layer_norm_epsilon": 1e-06,
|
| 18 |
+
"model_type": "t5",
|
| 19 |
+
"n_positions": 512,
|
| 20 |
+
"num_decoder_layers": 12,
|
| 21 |
+
"num_heads": 12,
|
| 22 |
+
"num_layers": 12,
|
| 23 |
+
"output_past": true,
|
| 24 |
+
"pad_token_id": 0,
|
| 25 |
+
"relative_attention_max_distance": 128,
|
| 26 |
+
"relative_attention_num_buckets": 32,
|
| 27 |
+
"task_specific_params": {
|
| 28 |
+
"summarization": {
|
| 29 |
+
"early_stopping": true,
|
| 30 |
+
"length_penalty": 2.0,
|
| 31 |
+
"max_length": 200,
|
| 32 |
+
"min_length": 30,
|
| 33 |
+
"no_repeat_ngram_size": 3,
|
| 34 |
+
"num_beams": 4,
|
| 35 |
+
"prefix": "summarize: "
|
| 36 |
+
},
|
| 37 |
+
"translation_en_to_de": {
|
| 38 |
+
"early_stopping": true,
|
| 39 |
+
"max_length": 300,
|
| 40 |
+
"num_beams": 4,
|
| 41 |
+
"prefix": "translate English to German: "
|
| 42 |
+
},
|
| 43 |
+
"translation_en_to_fr": {
|
| 44 |
+
"early_stopping": true,
|
| 45 |
+
"max_length": 300,
|
| 46 |
+
"num_beams": 4,
|
| 47 |
+
"prefix": "translate English to French: "
|
| 48 |
+
},
|
| 49 |
+
"translation_en_to_ro": {
|
| 50 |
+
"early_stopping": true,
|
| 51 |
+
"max_length": 300,
|
| 52 |
+
"num_beams": 4,
|
| 53 |
+
"prefix": "translate English to Romanian: "
|
| 54 |
+
}
|
| 55 |
+
},
|
| 56 |
+
"torch_dtype": "float32",
|
| 57 |
+
"transformers_version": "4.53.2",
|
| 58 |
+
"use_cache": true,
|
| 59 |
+
"vocab_size": 32128
|
| 60 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.53.2"
|
| 7 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1179071893c012f0ce2e0ef4559fb593df541240876131a940d8cd0c25b4714f
|
| 3 |
+
size 891644712
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:adb55039dbd0359f249f63ccf6125e32e50860a50691e23723966a95b1239581
|
| 3 |
+
size 1783444794
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c21b84d59605d01072d1bfd057b994626bcfcfd04f6f039f7ef79ba82de59bd7
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfe2ada38a28d1c9deff408ded137027e21df99e4a3d4de7cc8c4dbd7023a8f5
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2246 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 200,
|
| 7 |
+
"global_step": 2925,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03076923076923077,
|
| 14 |
+
"grad_norm": 24.726886749267578,
|
| 15 |
+
"learning_rate": 1.8e-06,
|
| 16 |
+
"loss": 11.4221,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06153846153846154,
|
| 21 |
+
"grad_norm": 17.795185089111328,
|
| 22 |
+
"learning_rate": 3.8e-06,
|
| 23 |
+
"loss": 10.4941,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09230769230769231,
|
| 28 |
+
"grad_norm": 14.8720703125,
|
| 29 |
+
"learning_rate": 5.8e-06,
|
| 30 |
+
"loss": 10.943,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.12307692307692308,
|
| 35 |
+
"grad_norm": 26.370025634765625,
|
| 36 |
+
"learning_rate": 7.8e-06,
|
| 37 |
+
"loss": 9.7676,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.15384615384615385,
|
| 42 |
+
"grad_norm": 25.46526527404785,
|
| 43 |
+
"learning_rate": 9.800000000000001e-06,
|
| 44 |
+
"loss": 8.0724,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.18461538461538463,
|
| 49 |
+
"grad_norm": 31.001745223999023,
|
| 50 |
+
"learning_rate": 1.18e-05,
|
| 51 |
+
"loss": 6.9668,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2153846153846154,
|
| 56 |
+
"grad_norm": 9.090025901794434,
|
| 57 |
+
"learning_rate": 1.3800000000000002e-05,
|
| 58 |
+
"loss": 5.0298,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.24615384615384617,
|
| 63 |
+
"grad_norm": 4.529256820678711,
|
| 64 |
+
"learning_rate": 1.58e-05,
|
| 65 |
+
"loss": 3.5286,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.27692307692307694,
|
| 70 |
+
"grad_norm": 2.912689685821533,
|
| 71 |
+
"learning_rate": 1.78e-05,
|
| 72 |
+
"loss": 2.9783,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3076923076923077,
|
| 77 |
+
"grad_norm": 2.04130220413208,
|
| 78 |
+
"learning_rate": 1.9800000000000004e-05,
|
| 79 |
+
"loss": 2.5693,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3384615384615385,
|
| 84 |
+
"grad_norm": 1.7645025253295898,
|
| 85 |
+
"learning_rate": 2.18e-05,
|
| 86 |
+
"loss": 2.3032,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.36923076923076925,
|
| 91 |
+
"grad_norm": 1.4153923988342285,
|
| 92 |
+
"learning_rate": 2.38e-05,
|
| 93 |
+
"loss": 2.1182,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4,
|
| 98 |
+
"grad_norm": 2.1566500663757324,
|
| 99 |
+
"learning_rate": 2.58e-05,
|
| 100 |
+
"loss": 2.2848,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4307692307692308,
|
| 105 |
+
"grad_norm": 1.7678470611572266,
|
| 106 |
+
"learning_rate": 2.7800000000000005e-05,
|
| 107 |
+
"loss": 2.2928,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.46153846153846156,
|
| 112 |
+
"grad_norm": 1.717806100845337,
|
| 113 |
+
"learning_rate": 2.98e-05,
|
| 114 |
+
"loss": 2.0866,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.49230769230769234,
|
| 119 |
+
"grad_norm": 1.8488136529922485,
|
| 120 |
+
"learning_rate": 3.18e-05,
|
| 121 |
+
"loss": 2.2323,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5230769230769231,
|
| 126 |
+
"grad_norm": 1.8968263864517212,
|
| 127 |
+
"learning_rate": 3.38e-05,
|
| 128 |
+
"loss": 2.2183,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5538461538461539,
|
| 133 |
+
"grad_norm": 1.398100733757019,
|
| 134 |
+
"learning_rate": 3.58e-05,
|
| 135 |
+
"loss": 1.8395,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.5846153846153846,
|
| 140 |
+
"grad_norm": 1.5084631443023682,
|
| 141 |
+
"learning_rate": 3.7800000000000004e-05,
|
| 142 |
+
"loss": 1.9694,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6153846153846154,
|
| 147 |
+
"grad_norm": 1.1019172668457031,
|
| 148 |
+
"learning_rate": 3.9800000000000005e-05,
|
| 149 |
+
"loss": 1.9702,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6153846153846154,
|
| 154 |
+
"eval_gen_len": 186.9088,
|
| 155 |
+
"eval_loss": 1.8701356649398804,
|
| 156 |
+
"eval_rouge1": 0.5735,
|
| 157 |
+
"eval_rouge2": 0.2231,
|
| 158 |
+
"eval_rougeL": 0.4346,
|
| 159 |
+
"eval_runtime": 33.3119,
|
| 160 |
+
"eval_samples_per_second": 8.225,
|
| 161 |
+
"eval_steps_per_second": 2.071,
|
| 162 |
+
"step": 200
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.6461538461538462,
|
| 166 |
+
"grad_norm": 1.5155857801437378,
|
| 167 |
+
"learning_rate": 4.18e-05,
|
| 168 |
+
"loss": 2.0705,
|
| 169 |
+
"step": 210
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.676923076923077,
|
| 173 |
+
"grad_norm": 1.3338611125946045,
|
| 174 |
+
"learning_rate": 4.38e-05,
|
| 175 |
+
"loss": 2.051,
|
| 176 |
+
"step": 220
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.7076923076923077,
|
| 180 |
+
"grad_norm": 1.2012193202972412,
|
| 181 |
+
"learning_rate": 4.58e-05,
|
| 182 |
+
"loss": 1.9834,
|
| 183 |
+
"step": 230
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.7384615384615385,
|
| 187 |
+
"grad_norm": 1.527007818222046,
|
| 188 |
+
"learning_rate": 4.78e-05,
|
| 189 |
+
"loss": 2.0451,
|
| 190 |
+
"step": 240
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.7692307692307693,
|
| 194 |
+
"grad_norm": 1.2146987915039062,
|
| 195 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 196 |
+
"loss": 1.974,
|
| 197 |
+
"step": 250
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.8,
|
| 201 |
+
"grad_norm": 1.7576699256896973,
|
| 202 |
+
"learning_rate": 5.1800000000000005e-05,
|
| 203 |
+
"loss": 2.0347,
|
| 204 |
+
"step": 260
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.8307692307692308,
|
| 208 |
+
"grad_norm": 1.17750084400177,
|
| 209 |
+
"learning_rate": 5.380000000000001e-05,
|
| 210 |
+
"loss": 1.9763,
|
| 211 |
+
"step": 270
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.8615384615384616,
|
| 215 |
+
"grad_norm": 1.3267815113067627,
|
| 216 |
+
"learning_rate": 5.580000000000001e-05,
|
| 217 |
+
"loss": 1.842,
|
| 218 |
+
"step": 280
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.8923076923076924,
|
| 222 |
+
"grad_norm": 1.0520875453948975,
|
| 223 |
+
"learning_rate": 5.7799999999999995e-05,
|
| 224 |
+
"loss": 2.0525,
|
| 225 |
+
"step": 290
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.9230769230769231,
|
| 229 |
+
"grad_norm": 1.4701600074768066,
|
| 230 |
+
"learning_rate": 5.9800000000000003e-05,
|
| 231 |
+
"loss": 1.7418,
|
| 232 |
+
"step": 300
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.9538461538461539,
|
| 236 |
+
"grad_norm": 1.062267780303955,
|
| 237 |
+
"learning_rate": 6.18e-05,
|
| 238 |
+
"loss": 1.9685,
|
| 239 |
+
"step": 310
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.9846153846153847,
|
| 243 |
+
"grad_norm": 1.2724727392196655,
|
| 244 |
+
"learning_rate": 6.38e-05,
|
| 245 |
+
"loss": 1.7972,
|
| 246 |
+
"step": 320
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.0153846153846153,
|
| 250 |
+
"grad_norm": 1.2834393978118896,
|
| 251 |
+
"learning_rate": 6.58e-05,
|
| 252 |
+
"loss": 1.8395,
|
| 253 |
+
"step": 330
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.0461538461538462,
|
| 257 |
+
"grad_norm": 0.9657095074653625,
|
| 258 |
+
"learning_rate": 6.780000000000001e-05,
|
| 259 |
+
"loss": 1.836,
|
| 260 |
+
"step": 340
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.0769230769230769,
|
| 264 |
+
"grad_norm": 1.0390011072158813,
|
| 265 |
+
"learning_rate": 6.98e-05,
|
| 266 |
+
"loss": 1.9328,
|
| 267 |
+
"step": 350
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.1076923076923078,
|
| 271 |
+
"grad_norm": 1.2896322011947632,
|
| 272 |
+
"learning_rate": 7.18e-05,
|
| 273 |
+
"loss": 1.8227,
|
| 274 |
+
"step": 360
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.1384615384615384,
|
| 278 |
+
"grad_norm": 1.53290593624115,
|
| 279 |
+
"learning_rate": 7.38e-05,
|
| 280 |
+
"loss": 1.9214,
|
| 281 |
+
"step": 370
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.1692307692307693,
|
| 285 |
+
"grad_norm": 1.0145893096923828,
|
| 286 |
+
"learning_rate": 7.58e-05,
|
| 287 |
+
"loss": 1.8295,
|
| 288 |
+
"step": 380
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 1.2,
|
| 292 |
+
"grad_norm": 1.2127511501312256,
|
| 293 |
+
"learning_rate": 7.780000000000001e-05,
|
| 294 |
+
"loss": 1.7774,
|
| 295 |
+
"step": 390
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 1.2307692307692308,
|
| 299 |
+
"grad_norm": 1.1971853971481323,
|
| 300 |
+
"learning_rate": 7.98e-05,
|
| 301 |
+
"loss": 1.9926,
|
| 302 |
+
"step": 400
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.2307692307692308,
|
| 306 |
+
"eval_gen_len": 187.0146,
|
| 307 |
+
"eval_loss": 1.7412512302398682,
|
| 308 |
+
"eval_rouge1": 0.5961,
|
| 309 |
+
"eval_rouge2": 0.2459,
|
| 310 |
+
"eval_rougeL": 0.4577,
|
| 311 |
+
"eval_runtime": 33.1213,
|
| 312 |
+
"eval_samples_per_second": 8.273,
|
| 313 |
+
"eval_steps_per_second": 2.083,
|
| 314 |
+
"step": 400
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.2615384615384615,
|
| 318 |
+
"grad_norm": 1.107351303100586,
|
| 319 |
+
"learning_rate": 8.18e-05,
|
| 320 |
+
"loss": 1.8515,
|
| 321 |
+
"step": 410
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.2923076923076924,
|
| 325 |
+
"grad_norm": 1.1407504081726074,
|
| 326 |
+
"learning_rate": 8.38e-05,
|
| 327 |
+
"loss": 1.7011,
|
| 328 |
+
"step": 420
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 1.323076923076923,
|
| 332 |
+
"grad_norm": 1.418338656425476,
|
| 333 |
+
"learning_rate": 8.58e-05,
|
| 334 |
+
"loss": 1.6384,
|
| 335 |
+
"step": 430
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.353846153846154,
|
| 339 |
+
"grad_norm": 1.3715286254882812,
|
| 340 |
+
"learning_rate": 8.78e-05,
|
| 341 |
+
"loss": 1.8502,
|
| 342 |
+
"step": 440
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.3846153846153846,
|
| 346 |
+
"grad_norm": 0.9774390459060669,
|
| 347 |
+
"learning_rate": 8.98e-05,
|
| 348 |
+
"loss": 1.8264,
|
| 349 |
+
"step": 450
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 1.4153846153846155,
|
| 353 |
+
"grad_norm": 1.4778176546096802,
|
| 354 |
+
"learning_rate": 9.180000000000001e-05,
|
| 355 |
+
"loss": 1.694,
|
| 356 |
+
"step": 460
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.4461538461538461,
|
| 360 |
+
"grad_norm": 1.2721563577651978,
|
| 361 |
+
"learning_rate": 9.38e-05,
|
| 362 |
+
"loss": 1.8213,
|
| 363 |
+
"step": 470
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.476923076923077,
|
| 367 |
+
"grad_norm": 0.94813472032547,
|
| 368 |
+
"learning_rate": 9.58e-05,
|
| 369 |
+
"loss": 1.6636,
|
| 370 |
+
"step": 480
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 1.5076923076923077,
|
| 374 |
+
"grad_norm": 1.0905983448028564,
|
| 375 |
+
"learning_rate": 9.78e-05,
|
| 376 |
+
"loss": 1.7712,
|
| 377 |
+
"step": 490
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.5384615384615383,
|
| 381 |
+
"grad_norm": 1.1593286991119385,
|
| 382 |
+
"learning_rate": 9.98e-05,
|
| 383 |
+
"loss": 1.808,
|
| 384 |
+
"step": 500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.5692307692307692,
|
| 388 |
+
"grad_norm": 1.0593713521957397,
|
| 389 |
+
"learning_rate": 9.967272727272727e-05,
|
| 390 |
+
"loss": 1.82,
|
| 391 |
+
"step": 510
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.6,
|
| 395 |
+
"grad_norm": 0.941973865032196,
|
| 396 |
+
"learning_rate": 9.930909090909092e-05,
|
| 397 |
+
"loss": 1.7341,
|
| 398 |
+
"step": 520
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.6307692307692307,
|
| 402 |
+
"grad_norm": 0.80891352891922,
|
| 403 |
+
"learning_rate": 9.894545454545455e-05,
|
| 404 |
+
"loss": 1.6166,
|
| 405 |
+
"step": 530
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.6615384615384614,
|
| 409 |
+
"grad_norm": 1.0325396060943604,
|
| 410 |
+
"learning_rate": 9.858181818181819e-05,
|
| 411 |
+
"loss": 1.8333,
|
| 412 |
+
"step": 540
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 1.6923076923076923,
|
| 416 |
+
"grad_norm": 1.3042590618133545,
|
| 417 |
+
"learning_rate": 9.821818181818182e-05,
|
| 418 |
+
"loss": 1.6287,
|
| 419 |
+
"step": 550
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.7230769230769232,
|
| 423 |
+
"grad_norm": 1.475900650024414,
|
| 424 |
+
"learning_rate": 9.785454545454545e-05,
|
| 425 |
+
"loss": 1.6019,
|
| 426 |
+
"step": 560
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 1.7538461538461538,
|
| 430 |
+
"grad_norm": 1.1589939594268799,
|
| 431 |
+
"learning_rate": 9.74909090909091e-05,
|
| 432 |
+
"loss": 1.6904,
|
| 433 |
+
"step": 570
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 1.7846153846153845,
|
| 437 |
+
"grad_norm": 1.2714788913726807,
|
| 438 |
+
"learning_rate": 9.712727272727274e-05,
|
| 439 |
+
"loss": 1.7928,
|
| 440 |
+
"step": 580
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.8153846153846154,
|
| 444 |
+
"grad_norm": 1.2037074565887451,
|
| 445 |
+
"learning_rate": 9.676363636363637e-05,
|
| 446 |
+
"loss": 1.8325,
|
| 447 |
+
"step": 590
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.8461538461538463,
|
| 451 |
+
"grad_norm": 1.1115801334381104,
|
| 452 |
+
"learning_rate": 9.64e-05,
|
| 453 |
+
"loss": 1.7673,
|
| 454 |
+
"step": 600
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.8461538461538463,
|
| 458 |
+
"eval_gen_len": 187.0146,
|
| 459 |
+
"eval_loss": 1.69492506980896,
|
| 460 |
+
"eval_rouge1": 0.6004,
|
| 461 |
+
"eval_rouge2": 0.2505,
|
| 462 |
+
"eval_rougeL": 0.4658,
|
| 463 |
+
"eval_runtime": 33.2205,
|
| 464 |
+
"eval_samples_per_second": 8.248,
|
| 465 |
+
"eval_steps_per_second": 2.077,
|
| 466 |
+
"step": 600
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.876923076923077,
|
| 470 |
+
"grad_norm": 1.2008461952209473,
|
| 471 |
+
"learning_rate": 9.603636363636364e-05,
|
| 472 |
+
"loss": 1.7674,
|
| 473 |
+
"step": 610
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.9076923076923076,
|
| 477 |
+
"grad_norm": 1.1482900381088257,
|
| 478 |
+
"learning_rate": 9.567272727272729e-05,
|
| 479 |
+
"loss": 1.7932,
|
| 480 |
+
"step": 620
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.9384615384615385,
|
| 484 |
+
"grad_norm": 1.0144352912902832,
|
| 485 |
+
"learning_rate": 9.530909090909092e-05,
|
| 486 |
+
"loss": 1.6315,
|
| 487 |
+
"step": 630
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.9692307692307693,
|
| 491 |
+
"grad_norm": 0.9276631474494934,
|
| 492 |
+
"learning_rate": 9.494545454545455e-05,
|
| 493 |
+
"loss": 1.8373,
|
| 494 |
+
"step": 640
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.0,
|
| 498 |
+
"grad_norm": 1.0593888759613037,
|
| 499 |
+
"learning_rate": 9.458181818181819e-05,
|
| 500 |
+
"loss": 1.7967,
|
| 501 |
+
"step": 650
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.0307692307692307,
|
| 505 |
+
"grad_norm": 1.1406164169311523,
|
| 506 |
+
"learning_rate": 9.421818181818183e-05,
|
| 507 |
+
"loss": 1.7668,
|
| 508 |
+
"step": 660
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.0615384615384613,
|
| 512 |
+
"grad_norm": 0.9809508919715881,
|
| 513 |
+
"learning_rate": 9.385454545454546e-05,
|
| 514 |
+
"loss": 1.6602,
|
| 515 |
+
"step": 670
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.0923076923076924,
|
| 519 |
+
"grad_norm": 1.1698426008224487,
|
| 520 |
+
"learning_rate": 9.349090909090909e-05,
|
| 521 |
+
"loss": 1.6775,
|
| 522 |
+
"step": 680
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.123076923076923,
|
| 526 |
+
"grad_norm": 1.1539372205734253,
|
| 527 |
+
"learning_rate": 9.312727272727274e-05,
|
| 528 |
+
"loss": 1.533,
|
| 529 |
+
"step": 690
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.1538461538461537,
|
| 533 |
+
"grad_norm": 1.0783981084823608,
|
| 534 |
+
"learning_rate": 9.276363636363637e-05,
|
| 535 |
+
"loss": 1.5243,
|
| 536 |
+
"step": 700
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.184615384615385,
|
| 540 |
+
"grad_norm": 1.262705683708191,
|
| 541 |
+
"learning_rate": 9.240000000000001e-05,
|
| 542 |
+
"loss": 1.6625,
|
| 543 |
+
"step": 710
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.2153846153846155,
|
| 547 |
+
"grad_norm": 1.1545718908309937,
|
| 548 |
+
"learning_rate": 9.203636363636364e-05,
|
| 549 |
+
"loss": 1.9172,
|
| 550 |
+
"step": 720
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.246153846153846,
|
| 554 |
+
"grad_norm": 0.9896947741508484,
|
| 555 |
+
"learning_rate": 9.167272727272728e-05,
|
| 556 |
+
"loss": 1.5449,
|
| 557 |
+
"step": 730
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.276923076923077,
|
| 561 |
+
"grad_norm": 1.063262701034546,
|
| 562 |
+
"learning_rate": 9.130909090909091e-05,
|
| 563 |
+
"loss": 1.5318,
|
| 564 |
+
"step": 740
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.3076923076923075,
|
| 568 |
+
"grad_norm": 1.0846728086471558,
|
| 569 |
+
"learning_rate": 9.094545454545454e-05,
|
| 570 |
+
"loss": 1.5875,
|
| 571 |
+
"step": 750
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.3384615384615386,
|
| 575 |
+
"grad_norm": 1.0437549352645874,
|
| 576 |
+
"learning_rate": 9.058181818181819e-05,
|
| 577 |
+
"loss": 1.5724,
|
| 578 |
+
"step": 760
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.3692307692307693,
|
| 582 |
+
"grad_norm": 1.050115942955017,
|
| 583 |
+
"learning_rate": 9.021818181818183e-05,
|
| 584 |
+
"loss": 1.662,
|
| 585 |
+
"step": 770
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.4,
|
| 589 |
+
"grad_norm": 0.9663347601890564,
|
| 590 |
+
"learning_rate": 8.985454545454546e-05,
|
| 591 |
+
"loss": 1.6284,
|
| 592 |
+
"step": 780
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.430769230769231,
|
| 596 |
+
"grad_norm": 1.1656932830810547,
|
| 597 |
+
"learning_rate": 8.949090909090909e-05,
|
| 598 |
+
"loss": 1.5995,
|
| 599 |
+
"step": 790
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.4615384615384617,
|
| 603 |
+
"grad_norm": 1.073716402053833,
|
| 604 |
+
"learning_rate": 8.912727272727273e-05,
|
| 605 |
+
"loss": 1.4811,
|
| 606 |
+
"step": 800
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.4615384615384617,
|
| 610 |
+
"eval_gen_len": 187.0146,
|
| 611 |
+
"eval_loss": 1.6769312620162964,
|
| 612 |
+
"eval_rouge1": 0.6042,
|
| 613 |
+
"eval_rouge2": 0.2561,
|
| 614 |
+
"eval_rougeL": 0.4686,
|
| 615 |
+
"eval_runtime": 33.5273,
|
| 616 |
+
"eval_samples_per_second": 8.172,
|
| 617 |
+
"eval_steps_per_second": 2.058,
|
| 618 |
+
"step": 800
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.4923076923076923,
|
| 622 |
+
"grad_norm": 0.9764583110809326,
|
| 623 |
+
"learning_rate": 8.876363636363638e-05,
|
| 624 |
+
"loss": 1.5478,
|
| 625 |
+
"step": 810
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.523076923076923,
|
| 629 |
+
"grad_norm": 0.9336417317390442,
|
| 630 |
+
"learning_rate": 8.840000000000001e-05,
|
| 631 |
+
"loss": 1.5138,
|
| 632 |
+
"step": 820
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.5538461538461537,
|
| 636 |
+
"grad_norm": 0.9714758992195129,
|
| 637 |
+
"learning_rate": 8.803636363636364e-05,
|
| 638 |
+
"loss": 1.5506,
|
| 639 |
+
"step": 830
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.5846153846153848,
|
| 643 |
+
"grad_norm": 0.9208464622497559,
|
| 644 |
+
"learning_rate": 8.767272727272727e-05,
|
| 645 |
+
"loss": 1.4944,
|
| 646 |
+
"step": 840
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.6153846153846154,
|
| 650 |
+
"grad_norm": 1.0252026319503784,
|
| 651 |
+
"learning_rate": 8.730909090909092e-05,
|
| 652 |
+
"loss": 1.6991,
|
| 653 |
+
"step": 850
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.646153846153846,
|
| 657 |
+
"grad_norm": 1.0464015007019043,
|
| 658 |
+
"learning_rate": 8.694545454545455e-05,
|
| 659 |
+
"loss": 1.679,
|
| 660 |
+
"step": 860
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.676923076923077,
|
| 664 |
+
"grad_norm": 1.3673149347305298,
|
| 665 |
+
"learning_rate": 8.658181818181818e-05,
|
| 666 |
+
"loss": 1.5021,
|
| 667 |
+
"step": 870
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.707692307692308,
|
| 671 |
+
"grad_norm": 1.1350778341293335,
|
| 672 |
+
"learning_rate": 8.621818181818181e-05,
|
| 673 |
+
"loss": 1.5898,
|
| 674 |
+
"step": 880
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.7384615384615385,
|
| 678 |
+
"grad_norm": 0.9916401505470276,
|
| 679 |
+
"learning_rate": 8.585454545454546e-05,
|
| 680 |
+
"loss": 1.6542,
|
| 681 |
+
"step": 890
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.769230769230769,
|
| 685 |
+
"grad_norm": 0.9967766404151917,
|
| 686 |
+
"learning_rate": 8.54909090909091e-05,
|
| 687 |
+
"loss": 1.7056,
|
| 688 |
+
"step": 900
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.8,
|
| 692 |
+
"grad_norm": 1.2031991481781006,
|
| 693 |
+
"learning_rate": 8.512727272727273e-05,
|
| 694 |
+
"loss": 1.6856,
|
| 695 |
+
"step": 910
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.830769230769231,
|
| 699 |
+
"grad_norm": 1.0159794092178345,
|
| 700 |
+
"learning_rate": 8.476363636363636e-05,
|
| 701 |
+
"loss": 1.5293,
|
| 702 |
+
"step": 920
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.8615384615384616,
|
| 706 |
+
"grad_norm": 1.3572866916656494,
|
| 707 |
+
"learning_rate": 8.44e-05,
|
| 708 |
+
"loss": 1.6191,
|
| 709 |
+
"step": 930
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.8923076923076922,
|
| 713 |
+
"grad_norm": 1.2567291259765625,
|
| 714 |
+
"learning_rate": 8.403636363636364e-05,
|
| 715 |
+
"loss": 1.7504,
|
| 716 |
+
"step": 940
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.9230769230769234,
|
| 720 |
+
"grad_norm": 1.2280553579330444,
|
| 721 |
+
"learning_rate": 8.367272727272728e-05,
|
| 722 |
+
"loss": 1.6523,
|
| 723 |
+
"step": 950
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.953846153846154,
|
| 727 |
+
"grad_norm": 1.0409953594207764,
|
| 728 |
+
"learning_rate": 8.330909090909091e-05,
|
| 729 |
+
"loss": 1.5903,
|
| 730 |
+
"step": 960
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.9846153846153847,
|
| 734 |
+
"grad_norm": 1.10386061668396,
|
| 735 |
+
"learning_rate": 8.294545454545455e-05,
|
| 736 |
+
"loss": 1.5235,
|
| 737 |
+
"step": 970
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.0153846153846153,
|
| 741 |
+
"grad_norm": 1.0341882705688477,
|
| 742 |
+
"learning_rate": 8.258181818181818e-05,
|
| 743 |
+
"loss": 1.6025,
|
| 744 |
+
"step": 980
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.046153846153846,
|
| 748 |
+
"grad_norm": 1.3020343780517578,
|
| 749 |
+
"learning_rate": 8.221818181818183e-05,
|
| 750 |
+
"loss": 1.4696,
|
| 751 |
+
"step": 990
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.076923076923077,
|
| 755 |
+
"grad_norm": 1.104643702507019,
|
| 756 |
+
"learning_rate": 8.185454545454546e-05,
|
| 757 |
+
"loss": 1.4009,
|
| 758 |
+
"step": 1000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.076923076923077,
|
| 762 |
+
"eval_gen_len": 187.0146,
|
| 763 |
+
"eval_loss": 1.6721168756484985,
|
| 764 |
+
"eval_rouge1": 0.6044,
|
| 765 |
+
"eval_rouge2": 0.2558,
|
| 766 |
+
"eval_rougeL": 0.4692,
|
| 767 |
+
"eval_runtime": 34.4765,
|
| 768 |
+
"eval_samples_per_second": 7.947,
|
| 769 |
+
"eval_steps_per_second": 2.001,
|
| 770 |
+
"step": 1000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 3.1076923076923078,
|
| 774 |
+
"grad_norm": 1.1327263116836548,
|
| 775 |
+
"learning_rate": 8.14909090909091e-05,
|
| 776 |
+
"loss": 1.5893,
|
| 777 |
+
"step": 1010
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 3.1384615384615384,
|
| 781 |
+
"grad_norm": 1.168095350265503,
|
| 782 |
+
"learning_rate": 8.112727272727273e-05,
|
| 783 |
+
"loss": 1.4248,
|
| 784 |
+
"step": 1020
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 3.169230769230769,
|
| 788 |
+
"grad_norm": 0.9978489279747009,
|
| 789 |
+
"learning_rate": 8.076363636363636e-05,
|
| 790 |
+
"loss": 1.5407,
|
| 791 |
+
"step": 1030
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 3.2,
|
| 795 |
+
"grad_norm": 1.0370062589645386,
|
| 796 |
+
"learning_rate": 8.04e-05,
|
| 797 |
+
"loss": 1.4867,
|
| 798 |
+
"step": 1040
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 3.230769230769231,
|
| 802 |
+
"grad_norm": 0.9647369384765625,
|
| 803 |
+
"learning_rate": 8.003636363636365e-05,
|
| 804 |
+
"loss": 1.4806,
|
| 805 |
+
"step": 1050
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 3.2615384615384615,
|
| 809 |
+
"grad_norm": 1.3316948413848877,
|
| 810 |
+
"learning_rate": 7.967272727272728e-05,
|
| 811 |
+
"loss": 1.4612,
|
| 812 |
+
"step": 1060
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 3.292307692307692,
|
| 816 |
+
"grad_norm": 1.37971830368042,
|
| 817 |
+
"learning_rate": 7.93090909090909e-05,
|
| 818 |
+
"loss": 1.5745,
|
| 819 |
+
"step": 1070
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 3.3230769230769233,
|
| 823 |
+
"grad_norm": 1.1220242977142334,
|
| 824 |
+
"learning_rate": 7.894545454545455e-05,
|
| 825 |
+
"loss": 1.3228,
|
| 826 |
+
"step": 1080
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 3.353846153846154,
|
| 830 |
+
"grad_norm": 1.0595531463623047,
|
| 831 |
+
"learning_rate": 7.85818181818182e-05,
|
| 832 |
+
"loss": 1.4618,
|
| 833 |
+
"step": 1090
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 3.3846153846153846,
|
| 837 |
+
"grad_norm": 1.3739666938781738,
|
| 838 |
+
"learning_rate": 7.821818181818182e-05,
|
| 839 |
+
"loss": 1.4973,
|
| 840 |
+
"step": 1100
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 3.4153846153846152,
|
| 844 |
+
"grad_norm": 1.2643866539001465,
|
| 845 |
+
"learning_rate": 7.785454545454545e-05,
|
| 846 |
+
"loss": 1.533,
|
| 847 |
+
"step": 1110
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 3.4461538461538463,
|
| 851 |
+
"grad_norm": 1.232230544090271,
|
| 852 |
+
"learning_rate": 7.74909090909091e-05,
|
| 853 |
+
"loss": 1.4867,
|
| 854 |
+
"step": 1120
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 3.476923076923077,
|
| 858 |
+
"grad_norm": 0.9712868332862854,
|
| 859 |
+
"learning_rate": 7.712727272727273e-05,
|
| 860 |
+
"loss": 1.4916,
|
| 861 |
+
"step": 1130
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 3.5076923076923077,
|
| 865 |
+
"grad_norm": 1.3414063453674316,
|
| 866 |
+
"learning_rate": 7.676363636363637e-05,
|
| 867 |
+
"loss": 1.5992,
|
| 868 |
+
"step": 1140
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 3.5384615384615383,
|
| 872 |
+
"grad_norm": 1.0222588777542114,
|
| 873 |
+
"learning_rate": 7.64e-05,
|
| 874 |
+
"loss": 1.5378,
|
| 875 |
+
"step": 1150
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 3.569230769230769,
|
| 879 |
+
"grad_norm": 1.1905276775360107,
|
| 880 |
+
"learning_rate": 7.603636363636364e-05,
|
| 881 |
+
"loss": 1.4324,
|
| 882 |
+
"step": 1160
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 3.6,
|
| 886 |
+
"grad_norm": 1.2846956253051758,
|
| 887 |
+
"learning_rate": 7.567272727272727e-05,
|
| 888 |
+
"loss": 1.4927,
|
| 889 |
+
"step": 1170
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 3.6307692307692307,
|
| 893 |
+
"grad_norm": 1.165310025215149,
|
| 894 |
+
"learning_rate": 7.530909090909092e-05,
|
| 895 |
+
"loss": 1.4277,
|
| 896 |
+
"step": 1180
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 3.6615384615384614,
|
| 900 |
+
"grad_norm": 1.063883900642395,
|
| 901 |
+
"learning_rate": 7.494545454545455e-05,
|
| 902 |
+
"loss": 1.5226,
|
| 903 |
+
"step": 1190
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 3.6923076923076925,
|
| 907 |
+
"grad_norm": 1.3430577516555786,
|
| 908 |
+
"learning_rate": 7.458181818181819e-05,
|
| 909 |
+
"loss": 1.5315,
|
| 910 |
+
"step": 1200
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 3.6923076923076925,
|
| 914 |
+
"eval_gen_len": 187.0146,
|
| 915 |
+
"eval_loss": 1.6592342853546143,
|
| 916 |
+
"eval_rouge1": 0.6057,
|
| 917 |
+
"eval_rouge2": 0.2572,
|
| 918 |
+
"eval_rougeL": 0.4709,
|
| 919 |
+
"eval_runtime": 33.3732,
|
| 920 |
+
"eval_samples_per_second": 8.21,
|
| 921 |
+
"eval_steps_per_second": 2.068,
|
| 922 |
+
"step": 1200
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 3.723076923076923,
|
| 926 |
+
"grad_norm": 1.2028673887252808,
|
| 927 |
+
"learning_rate": 7.421818181818182e-05,
|
| 928 |
+
"loss": 1.5135,
|
| 929 |
+
"step": 1210
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.753846153846154,
|
| 933 |
+
"grad_norm": 0.9091282486915588,
|
| 934 |
+
"learning_rate": 7.385454545454545e-05,
|
| 935 |
+
"loss": 1.3876,
|
| 936 |
+
"step": 1220
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.7846153846153845,
|
| 940 |
+
"grad_norm": 0.9549902677536011,
|
| 941 |
+
"learning_rate": 7.34909090909091e-05,
|
| 942 |
+
"loss": 1.7408,
|
| 943 |
+
"step": 1230
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.815384615384615,
|
| 947 |
+
"grad_norm": 1.109423279762268,
|
| 948 |
+
"learning_rate": 7.312727272727274e-05,
|
| 949 |
+
"loss": 1.3897,
|
| 950 |
+
"step": 1240
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.8461538461538463,
|
| 954 |
+
"grad_norm": 1.1412984132766724,
|
| 955 |
+
"learning_rate": 7.276363636363637e-05,
|
| 956 |
+
"loss": 1.5034,
|
| 957 |
+
"step": 1250
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 3.876923076923077,
|
| 961 |
+
"grad_norm": 1.176283359527588,
|
| 962 |
+
"learning_rate": 7.24e-05,
|
| 963 |
+
"loss": 1.5426,
|
| 964 |
+
"step": 1260
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 3.9076923076923076,
|
| 968 |
+
"grad_norm": 1.3558485507965088,
|
| 969 |
+
"learning_rate": 7.203636363636364e-05,
|
| 970 |
+
"loss": 1.5287,
|
| 971 |
+
"step": 1270
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.9384615384615387,
|
| 975 |
+
"grad_norm": 1.23688805103302,
|
| 976 |
+
"learning_rate": 7.167272727272729e-05,
|
| 977 |
+
"loss": 1.5794,
|
| 978 |
+
"step": 1280
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 3.9692307692307693,
|
| 982 |
+
"grad_norm": 1.248528242111206,
|
| 983 |
+
"learning_rate": 7.130909090909092e-05,
|
| 984 |
+
"loss": 1.5166,
|
| 985 |
+
"step": 1290
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.0,
|
| 989 |
+
"grad_norm": 0.8396208882331848,
|
| 990 |
+
"learning_rate": 7.094545454545455e-05,
|
| 991 |
+
"loss": 1.4648,
|
| 992 |
+
"step": 1300
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.030769230769231,
|
| 996 |
+
"grad_norm": 1.0756162405014038,
|
| 997 |
+
"learning_rate": 7.058181818181819e-05,
|
| 998 |
+
"loss": 1.4147,
|
| 999 |
+
"step": 1310
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.061538461538461,
|
| 1003 |
+
"grad_norm": 1.1731723546981812,
|
| 1004 |
+
"learning_rate": 7.021818181818182e-05,
|
| 1005 |
+
"loss": 1.3732,
|
| 1006 |
+
"step": 1320
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.092307692307692,
|
| 1010 |
+
"grad_norm": 1.2993149757385254,
|
| 1011 |
+
"learning_rate": 6.985454545454546e-05,
|
| 1012 |
+
"loss": 1.4118,
|
| 1013 |
+
"step": 1330
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.123076923076923,
|
| 1017 |
+
"grad_norm": 1.1648204326629639,
|
| 1018 |
+
"learning_rate": 6.949090909090909e-05,
|
| 1019 |
+
"loss": 1.4258,
|
| 1020 |
+
"step": 1340
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.153846153846154,
|
| 1024 |
+
"grad_norm": 1.1242045164108276,
|
| 1025 |
+
"learning_rate": 6.912727272727274e-05,
|
| 1026 |
+
"loss": 1.3598,
|
| 1027 |
+
"step": 1350
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.184615384615385,
|
| 1031 |
+
"grad_norm": 1.53397536277771,
|
| 1032 |
+
"learning_rate": 6.876363636363637e-05,
|
| 1033 |
+
"loss": 1.3431,
|
| 1034 |
+
"step": 1360
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.2153846153846155,
|
| 1038 |
+
"grad_norm": 1.3859331607818604,
|
| 1039 |
+
"learning_rate": 6.840000000000001e-05,
|
| 1040 |
+
"loss": 1.5014,
|
| 1041 |
+
"step": 1370
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.246153846153846,
|
| 1045 |
+
"grad_norm": 1.0821737051010132,
|
| 1046 |
+
"learning_rate": 6.803636363636364e-05,
|
| 1047 |
+
"loss": 1.2638,
|
| 1048 |
+
"step": 1380
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.276923076923077,
|
| 1052 |
+
"grad_norm": 1.2648742198944092,
|
| 1053 |
+
"learning_rate": 6.767272727272728e-05,
|
| 1054 |
+
"loss": 1.4567,
|
| 1055 |
+
"step": 1390
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.3076923076923075,
|
| 1059 |
+
"grad_norm": 1.1357372999191284,
|
| 1060 |
+
"learning_rate": 6.730909090909091e-05,
|
| 1061 |
+
"loss": 1.4706,
|
| 1062 |
+
"step": 1400
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.3076923076923075,
|
| 1066 |
+
"eval_gen_len": 187.0146,
|
| 1067 |
+
"eval_loss": 1.6691502332687378,
|
| 1068 |
+
"eval_rouge1": 0.6061,
|
| 1069 |
+
"eval_rouge2": 0.2593,
|
| 1070 |
+
"eval_rougeL": 0.4719,
|
| 1071 |
+
"eval_runtime": 33.439,
|
| 1072 |
+
"eval_samples_per_second": 8.194,
|
| 1073 |
+
"eval_steps_per_second": 2.063,
|
| 1074 |
+
"step": 1400
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.338461538461538,
|
| 1078 |
+
"grad_norm": 1.4324264526367188,
|
| 1079 |
+
"learning_rate": 6.694545454545454e-05,
|
| 1080 |
+
"loss": 1.4032,
|
| 1081 |
+
"step": 1410
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.36923076923077,
|
| 1085 |
+
"grad_norm": 1.3646095991134644,
|
| 1086 |
+
"learning_rate": 6.658181818181819e-05,
|
| 1087 |
+
"loss": 1.3943,
|
| 1088 |
+
"step": 1420
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 4.4,
|
| 1092 |
+
"grad_norm": 0.9991398453712463,
|
| 1093 |
+
"learning_rate": 6.621818181818183e-05,
|
| 1094 |
+
"loss": 1.5292,
|
| 1095 |
+
"step": 1430
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 4.430769230769231,
|
| 1099 |
+
"grad_norm": 1.1873986721038818,
|
| 1100 |
+
"learning_rate": 6.585454545454546e-05,
|
| 1101 |
+
"loss": 1.4813,
|
| 1102 |
+
"step": 1440
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.461538461538462,
|
| 1106 |
+
"grad_norm": 1.0080267190933228,
|
| 1107 |
+
"learning_rate": 6.549090909090909e-05,
|
| 1108 |
+
"loss": 1.4951,
|
| 1109 |
+
"step": 1450
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 4.492307692307692,
|
| 1113 |
+
"grad_norm": 1.5542734861373901,
|
| 1114 |
+
"learning_rate": 6.512727272727272e-05,
|
| 1115 |
+
"loss": 1.5603,
|
| 1116 |
+
"step": 1460
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 4.523076923076923,
|
| 1120 |
+
"grad_norm": 1.2610498666763306,
|
| 1121 |
+
"learning_rate": 6.476363636363638e-05,
|
| 1122 |
+
"loss": 1.3286,
|
| 1123 |
+
"step": 1470
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 4.553846153846154,
|
| 1127 |
+
"grad_norm": 1.0882760286331177,
|
| 1128 |
+
"learning_rate": 6.440000000000001e-05,
|
| 1129 |
+
"loss": 1.309,
|
| 1130 |
+
"step": 1480
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 4.584615384615384,
|
| 1134 |
+
"grad_norm": 1.1589834690093994,
|
| 1135 |
+
"learning_rate": 6.403636363636364e-05,
|
| 1136 |
+
"loss": 1.414,
|
| 1137 |
+
"step": 1490
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.615384615384615,
|
| 1141 |
+
"grad_norm": 1.3731642961502075,
|
| 1142 |
+
"learning_rate": 6.367272727272727e-05,
|
| 1143 |
+
"loss": 1.4529,
|
| 1144 |
+
"step": 1500
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 4.6461538461538465,
|
| 1148 |
+
"grad_norm": 1.3053221702575684,
|
| 1149 |
+
"learning_rate": 6.330909090909091e-05,
|
| 1150 |
+
"loss": 1.4623,
|
| 1151 |
+
"step": 1510
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 4.676923076923077,
|
| 1155 |
+
"grad_norm": 1.2154396772384644,
|
| 1156 |
+
"learning_rate": 6.294545454545455e-05,
|
| 1157 |
+
"loss": 1.4766,
|
| 1158 |
+
"step": 1520
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 4.707692307692308,
|
| 1162 |
+
"grad_norm": 1.0947812795639038,
|
| 1163 |
+
"learning_rate": 6.258181818181818e-05,
|
| 1164 |
+
"loss": 1.3212,
|
| 1165 |
+
"step": 1530
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 4.7384615384615385,
|
| 1169 |
+
"grad_norm": 1.005462646484375,
|
| 1170 |
+
"learning_rate": 6.221818181818181e-05,
|
| 1171 |
+
"loss": 1.3956,
|
| 1172 |
+
"step": 1540
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.769230769230769,
|
| 1176 |
+
"grad_norm": 1.196108341217041,
|
| 1177 |
+
"learning_rate": 6.185454545454546e-05,
|
| 1178 |
+
"loss": 1.4048,
|
| 1179 |
+
"step": 1550
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 4.8,
|
| 1183 |
+
"grad_norm": 1.355747103691101,
|
| 1184 |
+
"learning_rate": 6.14909090909091e-05,
|
| 1185 |
+
"loss": 1.3474,
|
| 1186 |
+
"step": 1560
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 4.8307692307692305,
|
| 1190 |
+
"grad_norm": 1.177310585975647,
|
| 1191 |
+
"learning_rate": 6.112727272727273e-05,
|
| 1192 |
+
"loss": 1.3038,
|
| 1193 |
+
"step": 1570
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 4.861538461538462,
|
| 1197 |
+
"grad_norm": 1.273474097251892,
|
| 1198 |
+
"learning_rate": 6.076363636363637e-05,
|
| 1199 |
+
"loss": 1.3858,
|
| 1200 |
+
"step": 1580
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 4.892307692307693,
|
| 1204 |
+
"grad_norm": 1.2601613998413086,
|
| 1205 |
+
"learning_rate": 6.04e-05,
|
| 1206 |
+
"loss": 1.2742,
|
| 1207 |
+
"step": 1590
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 4.923076923076923,
|
| 1211 |
+
"grad_norm": 1.052040696144104,
|
| 1212 |
+
"learning_rate": 6.0036363636363634e-05,
|
| 1213 |
+
"loss": 1.5551,
|
| 1214 |
+
"step": 1600
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 4.923076923076923,
|
| 1218 |
+
"eval_gen_len": 187.0146,
|
| 1219 |
+
"eval_loss": 1.6678508520126343,
|
| 1220 |
+
"eval_rouge1": 0.6061,
|
| 1221 |
+
"eval_rouge2": 0.2582,
|
| 1222 |
+
"eval_rougeL": 0.4724,
|
| 1223 |
+
"eval_runtime": 33.608,
|
| 1224 |
+
"eval_samples_per_second": 8.153,
|
| 1225 |
+
"eval_steps_per_second": 2.053,
|
| 1226 |
+
"step": 1600
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 4.953846153846154,
|
| 1230 |
+
"grad_norm": 1.4437050819396973,
|
| 1231 |
+
"learning_rate": 5.967272727272728e-05,
|
| 1232 |
+
"loss": 1.3613,
|
| 1233 |
+
"step": 1610
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 4.984615384615385,
|
| 1237 |
+
"grad_norm": 1.398398518562317,
|
| 1238 |
+
"learning_rate": 5.9309090909090915e-05,
|
| 1239 |
+
"loss": 1.3861,
|
| 1240 |
+
"step": 1620
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 5.015384615384615,
|
| 1244 |
+
"grad_norm": 1.1901689767837524,
|
| 1245 |
+
"learning_rate": 5.894545454545455e-05,
|
| 1246 |
+
"loss": 1.4525,
|
| 1247 |
+
"step": 1630
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 5.046153846153846,
|
| 1251 |
+
"grad_norm": 1.2631349563598633,
|
| 1252 |
+
"learning_rate": 5.858181818181818e-05,
|
| 1253 |
+
"loss": 1.3698,
|
| 1254 |
+
"step": 1640
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 5.076923076923077,
|
| 1258 |
+
"grad_norm": 1.3628337383270264,
|
| 1259 |
+
"learning_rate": 5.821818181818182e-05,
|
| 1260 |
+
"loss": 1.1917,
|
| 1261 |
+
"step": 1650
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 5.107692307692307,
|
| 1265 |
+
"grad_norm": 1.2330440282821655,
|
| 1266 |
+
"learning_rate": 5.785454545454546e-05,
|
| 1267 |
+
"loss": 1.3692,
|
| 1268 |
+
"step": 1660
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 5.138461538461539,
|
| 1272 |
+
"grad_norm": 1.069877028465271,
|
| 1273 |
+
"learning_rate": 5.74909090909091e-05,
|
| 1274 |
+
"loss": 1.2439,
|
| 1275 |
+
"step": 1670
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 5.1692307692307695,
|
| 1279 |
+
"grad_norm": 1.1694751977920532,
|
| 1280 |
+
"learning_rate": 5.712727272727273e-05,
|
| 1281 |
+
"loss": 1.1194,
|
| 1282 |
+
"step": 1680
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 5.2,
|
| 1286 |
+
"grad_norm": 1.424668312072754,
|
| 1287 |
+
"learning_rate": 5.6763636363636365e-05,
|
| 1288 |
+
"loss": 1.456,
|
| 1289 |
+
"step": 1690
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 5.230769230769231,
|
| 1293 |
+
"grad_norm": 1.1166226863861084,
|
| 1294 |
+
"learning_rate": 5.6399999999999995e-05,
|
| 1295 |
+
"loss": 1.3011,
|
| 1296 |
+
"step": 1700
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 5.2615384615384615,
|
| 1300 |
+
"grad_norm": 1.196712851524353,
|
| 1301 |
+
"learning_rate": 5.6036363636363646e-05,
|
| 1302 |
+
"loss": 1.3272,
|
| 1303 |
+
"step": 1710
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 5.292307692307692,
|
| 1307 |
+
"grad_norm": 1.105592131614685,
|
| 1308 |
+
"learning_rate": 5.5672727272727276e-05,
|
| 1309 |
+
"loss": 1.3831,
|
| 1310 |
+
"step": 1720
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 5.323076923076923,
|
| 1314 |
+
"grad_norm": 1.3789408206939697,
|
| 1315 |
+
"learning_rate": 5.530909090909091e-05,
|
| 1316 |
+
"loss": 1.3924,
|
| 1317 |
+
"step": 1730
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 5.3538461538461535,
|
| 1321 |
+
"grad_norm": 1.1058343648910522,
|
| 1322 |
+
"learning_rate": 5.494545454545454e-05,
|
| 1323 |
+
"loss": 1.1278,
|
| 1324 |
+
"step": 1740
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 5.384615384615385,
|
| 1328 |
+
"grad_norm": 1.0470004081726074,
|
| 1329 |
+
"learning_rate": 5.458181818181819e-05,
|
| 1330 |
+
"loss": 1.2675,
|
| 1331 |
+
"step": 1750
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 5.415384615384616,
|
| 1335 |
+
"grad_norm": 1.5735996961593628,
|
| 1336 |
+
"learning_rate": 5.421818181818182e-05,
|
| 1337 |
+
"loss": 1.368,
|
| 1338 |
+
"step": 1760
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 5.446153846153846,
|
| 1342 |
+
"grad_norm": 1.053110122680664,
|
| 1343 |
+
"learning_rate": 5.385454545454546e-05,
|
| 1344 |
+
"loss": 1.4409,
|
| 1345 |
+
"step": 1770
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 5.476923076923077,
|
| 1349 |
+
"grad_norm": 1.2032136917114258,
|
| 1350 |
+
"learning_rate": 5.349090909090909e-05,
|
| 1351 |
+
"loss": 1.3919,
|
| 1352 |
+
"step": 1780
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 5.507692307692308,
|
| 1356 |
+
"grad_norm": 1.3398489952087402,
|
| 1357 |
+
"learning_rate": 5.3127272727272726e-05,
|
| 1358 |
+
"loss": 1.3891,
|
| 1359 |
+
"step": 1790
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 5.538461538461538,
|
| 1363 |
+
"grad_norm": 1.1674134731292725,
|
| 1364 |
+
"learning_rate": 5.276363636363637e-05,
|
| 1365 |
+
"loss": 1.4167,
|
| 1366 |
+
"step": 1800
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 5.538461538461538,
|
| 1370 |
+
"eval_gen_len": 187.0146,
|
| 1371 |
+
"eval_loss": 1.6820106506347656,
|
| 1372 |
+
"eval_rouge1": 0.6051,
|
| 1373 |
+
"eval_rouge2": 0.256,
|
| 1374 |
+
"eval_rougeL": 0.4705,
|
| 1375 |
+
"eval_runtime": 34.3835,
|
| 1376 |
+
"eval_samples_per_second": 7.969,
|
| 1377 |
+
"eval_steps_per_second": 2.007,
|
| 1378 |
+
"step": 1800
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 5.569230769230769,
|
| 1382 |
+
"grad_norm": 1.3142492771148682,
|
| 1383 |
+
"learning_rate": 5.2400000000000007e-05,
|
| 1384 |
+
"loss": 1.2869,
|
| 1385 |
+
"step": 1810
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 5.6,
|
| 1389 |
+
"grad_norm": 1.3149932622909546,
|
| 1390 |
+
"learning_rate": 5.2036363636363637e-05,
|
| 1391 |
+
"loss": 1.3653,
|
| 1392 |
+
"step": 1820
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 5.63076923076923,
|
| 1396 |
+
"grad_norm": 1.1139847040176392,
|
| 1397 |
+
"learning_rate": 5.167272727272727e-05,
|
| 1398 |
+
"loss": 1.3807,
|
| 1399 |
+
"step": 1830
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 5.661538461538462,
|
| 1403 |
+
"grad_norm": 1.302495002746582,
|
| 1404 |
+
"learning_rate": 5.130909090909091e-05,
|
| 1405 |
+
"loss": 1.3826,
|
| 1406 |
+
"step": 1840
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 5.6923076923076925,
|
| 1410 |
+
"grad_norm": 1.2617419958114624,
|
| 1411 |
+
"learning_rate": 5.0945454545454554e-05,
|
| 1412 |
+
"loss": 1.283,
|
| 1413 |
+
"step": 1850
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 5.723076923076923,
|
| 1417 |
+
"grad_norm": 1.252189040184021,
|
| 1418 |
+
"learning_rate": 5.0581818181818184e-05,
|
| 1419 |
+
"loss": 1.3425,
|
| 1420 |
+
"step": 1860
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 5.753846153846154,
|
| 1424 |
+
"grad_norm": 1.0823620557785034,
|
| 1425 |
+
"learning_rate": 5.021818181818182e-05,
|
| 1426 |
+
"loss": 1.4646,
|
| 1427 |
+
"step": 1870
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 5.7846153846153845,
|
| 1431 |
+
"grad_norm": 1.39573335647583,
|
| 1432 |
+
"learning_rate": 4.985454545454546e-05,
|
| 1433 |
+
"loss": 1.2957,
|
| 1434 |
+
"step": 1880
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 5.815384615384615,
|
| 1438 |
+
"grad_norm": 1.217499852180481,
|
| 1439 |
+
"learning_rate": 4.9490909090909094e-05,
|
| 1440 |
+
"loss": 1.3232,
|
| 1441 |
+
"step": 1890
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 5.846153846153846,
|
| 1445 |
+
"grad_norm": 1.3049825429916382,
|
| 1446 |
+
"learning_rate": 4.912727272727273e-05,
|
| 1447 |
+
"loss": 1.333,
|
| 1448 |
+
"step": 1900
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 5.876923076923077,
|
| 1452 |
+
"grad_norm": 1.265807032585144,
|
| 1453 |
+
"learning_rate": 4.876363636363637e-05,
|
| 1454 |
+
"loss": 1.2635,
|
| 1455 |
+
"step": 1910
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 5.907692307692308,
|
| 1459 |
+
"grad_norm": 1.4045813083648682,
|
| 1460 |
+
"learning_rate": 4.8400000000000004e-05,
|
| 1461 |
+
"loss": 1.3453,
|
| 1462 |
+
"step": 1920
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 5.938461538461539,
|
| 1466 |
+
"grad_norm": 1.4151256084442139,
|
| 1467 |
+
"learning_rate": 4.803636363636364e-05,
|
| 1468 |
+
"loss": 1.2274,
|
| 1469 |
+
"step": 1930
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 5.969230769230769,
|
| 1473 |
+
"grad_norm": 1.30918288230896,
|
| 1474 |
+
"learning_rate": 4.767272727272728e-05,
|
| 1475 |
+
"loss": 1.3835,
|
| 1476 |
+
"step": 1940
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.0,
|
| 1480 |
+
"grad_norm": 1.1394106149673462,
|
| 1481 |
+
"learning_rate": 4.7309090909090914e-05,
|
| 1482 |
+
"loss": 1.4773,
|
| 1483 |
+
"step": 1950
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 6.030769230769231,
|
| 1487 |
+
"grad_norm": 1.0991594791412354,
|
| 1488 |
+
"learning_rate": 4.694545454545455e-05,
|
| 1489 |
+
"loss": 1.1885,
|
| 1490 |
+
"step": 1960
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.061538461538461,
|
| 1494 |
+
"grad_norm": 1.2676807641983032,
|
| 1495 |
+
"learning_rate": 4.658181818181818e-05,
|
| 1496 |
+
"loss": 1.2931,
|
| 1497 |
+
"step": 1970
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.092307692307692,
|
| 1501 |
+
"grad_norm": 1.2844555377960205,
|
| 1502 |
+
"learning_rate": 4.6218181818181825e-05,
|
| 1503 |
+
"loss": 1.2758,
|
| 1504 |
+
"step": 1980
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 6.123076923076923,
|
| 1508 |
+
"grad_norm": 1.2364481687545776,
|
| 1509 |
+
"learning_rate": 4.5854545454545455e-05,
|
| 1510 |
+
"loss": 1.1871,
|
| 1511 |
+
"step": 1990
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 6.153846153846154,
|
| 1515 |
+
"grad_norm": 1.4142909049987793,
|
| 1516 |
+
"learning_rate": 4.54909090909091e-05,
|
| 1517 |
+
"loss": 1.2976,
|
| 1518 |
+
"step": 2000
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.153846153846154,
|
| 1522 |
+
"eval_gen_len": 187.0146,
|
| 1523 |
+
"eval_loss": 1.6960315704345703,
|
| 1524 |
+
"eval_rouge1": 0.6074,
|
| 1525 |
+
"eval_rouge2": 0.2564,
|
| 1526 |
+
"eval_rougeL": 0.4698,
|
| 1527 |
+
"eval_runtime": 33.4283,
|
| 1528 |
+
"eval_samples_per_second": 8.197,
|
| 1529 |
+
"eval_steps_per_second": 2.064,
|
| 1530 |
+
"step": 2000
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 6.184615384615385,
|
| 1534 |
+
"grad_norm": 1.234192132949829,
|
| 1535 |
+
"learning_rate": 4.512727272727273e-05,
|
| 1536 |
+
"loss": 1.3096,
|
| 1537 |
+
"step": 2010
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 6.2153846153846155,
|
| 1541 |
+
"grad_norm": 1.1222543716430664,
|
| 1542 |
+
"learning_rate": 4.4763636363636365e-05,
|
| 1543 |
+
"loss": 1.2937,
|
| 1544 |
+
"step": 2020
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 6.246153846153846,
|
| 1548 |
+
"grad_norm": 1.0864425897598267,
|
| 1549 |
+
"learning_rate": 4.44e-05,
|
| 1550 |
+
"loss": 1.278,
|
| 1551 |
+
"step": 2030
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 6.276923076923077,
|
| 1555 |
+
"grad_norm": 1.3431516885757446,
|
| 1556 |
+
"learning_rate": 4.403636363636364e-05,
|
| 1557 |
+
"loss": 1.2601,
|
| 1558 |
+
"step": 2040
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 6.3076923076923075,
|
| 1562 |
+
"grad_norm": 1.2604031562805176,
|
| 1563 |
+
"learning_rate": 4.3672727272727275e-05,
|
| 1564 |
+
"loss": 1.3587,
|
| 1565 |
+
"step": 2050
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 6.338461538461538,
|
| 1569 |
+
"grad_norm": 1.4237326383590698,
|
| 1570 |
+
"learning_rate": 4.330909090909091e-05,
|
| 1571 |
+
"loss": 1.3936,
|
| 1572 |
+
"step": 2060
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 6.36923076923077,
|
| 1576 |
+
"grad_norm": 1.2190274000167847,
|
| 1577 |
+
"learning_rate": 4.294545454545455e-05,
|
| 1578 |
+
"loss": 1.2805,
|
| 1579 |
+
"step": 2070
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 6.4,
|
| 1583 |
+
"grad_norm": 1.3023786544799805,
|
| 1584 |
+
"learning_rate": 4.2581818181818186e-05,
|
| 1585 |
+
"loss": 1.2676,
|
| 1586 |
+
"step": 2080
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 6.430769230769231,
|
| 1590 |
+
"grad_norm": 1.2170274257659912,
|
| 1591 |
+
"learning_rate": 4.2218181818181816e-05,
|
| 1592 |
+
"loss": 1.3696,
|
| 1593 |
+
"step": 2090
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 6.461538461538462,
|
| 1597 |
+
"grad_norm": 1.613784670829773,
|
| 1598 |
+
"learning_rate": 4.185454545454546e-05,
|
| 1599 |
+
"loss": 1.28,
|
| 1600 |
+
"step": 2100
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 6.492307692307692,
|
| 1604 |
+
"grad_norm": 1.2165530920028687,
|
| 1605 |
+
"learning_rate": 4.149090909090909e-05,
|
| 1606 |
+
"loss": 1.2933,
|
| 1607 |
+
"step": 2110
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 6.523076923076923,
|
| 1611 |
+
"grad_norm": 1.2213079929351807,
|
| 1612 |
+
"learning_rate": 4.112727272727273e-05,
|
| 1613 |
+
"loss": 1.2062,
|
| 1614 |
+
"step": 2120
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 6.553846153846154,
|
| 1618 |
+
"grad_norm": 1.5889174938201904,
|
| 1619 |
+
"learning_rate": 4.076363636363636e-05,
|
| 1620 |
+
"loss": 1.2481,
|
| 1621 |
+
"step": 2130
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 6.584615384615384,
|
| 1625 |
+
"grad_norm": 1.2638423442840576,
|
| 1626 |
+
"learning_rate": 4.0400000000000006e-05,
|
| 1627 |
+
"loss": 1.29,
|
| 1628 |
+
"step": 2140
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 6.615384615384615,
|
| 1632 |
+
"grad_norm": 1.0796576738357544,
|
| 1633 |
+
"learning_rate": 4.0036363636363636e-05,
|
| 1634 |
+
"loss": 1.1189,
|
| 1635 |
+
"step": 2150
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 6.6461538461538465,
|
| 1639 |
+
"grad_norm": 1.5910948514938354,
|
| 1640 |
+
"learning_rate": 3.967272727272727e-05,
|
| 1641 |
+
"loss": 1.2767,
|
| 1642 |
+
"step": 2160
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 6.676923076923077,
|
| 1646 |
+
"grad_norm": 1.3346668481826782,
|
| 1647 |
+
"learning_rate": 3.930909090909091e-05,
|
| 1648 |
+
"loss": 1.2212,
|
| 1649 |
+
"step": 2170
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 6.707692307692308,
|
| 1653 |
+
"grad_norm": 1.1277836561203003,
|
| 1654 |
+
"learning_rate": 3.8945454545454547e-05,
|
| 1655 |
+
"loss": 1.2392,
|
| 1656 |
+
"step": 2180
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 6.7384615384615385,
|
| 1660 |
+
"grad_norm": 1.1212108135223389,
|
| 1661 |
+
"learning_rate": 3.858181818181818e-05,
|
| 1662 |
+
"loss": 1.3223,
|
| 1663 |
+
"step": 2190
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 6.769230769230769,
|
| 1667 |
+
"grad_norm": 1.2175902128219604,
|
| 1668 |
+
"learning_rate": 3.821818181818182e-05,
|
| 1669 |
+
"loss": 1.3124,
|
| 1670 |
+
"step": 2200
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 6.769230769230769,
|
| 1674 |
+
"eval_gen_len": 187.0146,
|
| 1675 |
+
"eval_loss": 1.6920864582061768,
|
| 1676 |
+
"eval_rouge1": 0.606,
|
| 1677 |
+
"eval_rouge2": 0.2554,
|
| 1678 |
+
"eval_rougeL": 0.4694,
|
| 1679 |
+
"eval_runtime": 34.3853,
|
| 1680 |
+
"eval_samples_per_second": 7.969,
|
| 1681 |
+
"eval_steps_per_second": 2.007,
|
| 1682 |
+
"step": 2200
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 6.8,
|
| 1686 |
+
"grad_norm": 1.163404107093811,
|
| 1687 |
+
"learning_rate": 3.785454545454546e-05,
|
| 1688 |
+
"loss": 1.4557,
|
| 1689 |
+
"step": 2210
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 6.8307692307692305,
|
| 1693 |
+
"grad_norm": 1.5031542778015137,
|
| 1694 |
+
"learning_rate": 3.7490909090909094e-05,
|
| 1695 |
+
"loss": 1.1773,
|
| 1696 |
+
"step": 2220
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 6.861538461538462,
|
| 1700 |
+
"grad_norm": 1.0126748085021973,
|
| 1701 |
+
"learning_rate": 3.712727272727273e-05,
|
| 1702 |
+
"loss": 1.3922,
|
| 1703 |
+
"step": 2230
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 6.892307692307693,
|
| 1707 |
+
"grad_norm": 1.0135952234268188,
|
| 1708 |
+
"learning_rate": 3.676363636363637e-05,
|
| 1709 |
+
"loss": 1.2028,
|
| 1710 |
+
"step": 2240
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 6.923076923076923,
|
| 1714 |
+
"grad_norm": 1.16098153591156,
|
| 1715 |
+
"learning_rate": 3.6400000000000004e-05,
|
| 1716 |
+
"loss": 1.4111,
|
| 1717 |
+
"step": 2250
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 6.953846153846154,
|
| 1721 |
+
"grad_norm": 1.488234519958496,
|
| 1722 |
+
"learning_rate": 3.603636363636364e-05,
|
| 1723 |
+
"loss": 1.3131,
|
| 1724 |
+
"step": 2260
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.984615384615385,
|
| 1728 |
+
"grad_norm": 1.129989743232727,
|
| 1729 |
+
"learning_rate": 3.567272727272728e-05,
|
| 1730 |
+
"loss": 1.1701,
|
| 1731 |
+
"step": 2270
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.015384615384615,
|
| 1735 |
+
"grad_norm": 1.218468189239502,
|
| 1736 |
+
"learning_rate": 3.530909090909091e-05,
|
| 1737 |
+
"loss": 1.2604,
|
| 1738 |
+
"step": 2280
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.046153846153846,
|
| 1742 |
+
"grad_norm": 1.2339926958084106,
|
| 1743 |
+
"learning_rate": 3.494545454545455e-05,
|
| 1744 |
+
"loss": 1.0932,
|
| 1745 |
+
"step": 2290
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.076923076923077,
|
| 1749 |
+
"grad_norm": 1.4972765445709229,
|
| 1750 |
+
"learning_rate": 3.458181818181818e-05,
|
| 1751 |
+
"loss": 1.1137,
|
| 1752 |
+
"step": 2300
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.107692307692307,
|
| 1756 |
+
"grad_norm": 1.1884584426879883,
|
| 1757 |
+
"learning_rate": 3.4218181818181824e-05,
|
| 1758 |
+
"loss": 1.1522,
|
| 1759 |
+
"step": 2310
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.138461538461539,
|
| 1763 |
+
"grad_norm": 1.4934840202331543,
|
| 1764 |
+
"learning_rate": 3.3854545454545454e-05,
|
| 1765 |
+
"loss": 1.3121,
|
| 1766 |
+
"step": 2320
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.1692307692307695,
|
| 1770 |
+
"grad_norm": 1.1432678699493408,
|
| 1771 |
+
"learning_rate": 3.34909090909091e-05,
|
| 1772 |
+
"loss": 1.1549,
|
| 1773 |
+
"step": 2330
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.2,
|
| 1777 |
+
"grad_norm": 1.1708807945251465,
|
| 1778 |
+
"learning_rate": 3.312727272727273e-05,
|
| 1779 |
+
"loss": 1.1692,
|
| 1780 |
+
"step": 2340
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.230769230769231,
|
| 1784 |
+
"grad_norm": 1.3824517726898193,
|
| 1785 |
+
"learning_rate": 3.2763636363636365e-05,
|
| 1786 |
+
"loss": 1.208,
|
| 1787 |
+
"step": 2350
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.2615384615384615,
|
| 1791 |
+
"grad_norm": 1.1225407123565674,
|
| 1792 |
+
"learning_rate": 3.24e-05,
|
| 1793 |
+
"loss": 1.1542,
|
| 1794 |
+
"step": 2360
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.292307692307692,
|
| 1798 |
+
"grad_norm": 1.2445507049560547,
|
| 1799 |
+
"learning_rate": 3.203636363636364e-05,
|
| 1800 |
+
"loss": 1.2265,
|
| 1801 |
+
"step": 2370
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.323076923076923,
|
| 1805 |
+
"grad_norm": 1.256062626838684,
|
| 1806 |
+
"learning_rate": 3.1672727272727275e-05,
|
| 1807 |
+
"loss": 1.1822,
|
| 1808 |
+
"step": 2380
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.3538461538461535,
|
| 1812 |
+
"grad_norm": 1.3986501693725586,
|
| 1813 |
+
"learning_rate": 3.130909090909091e-05,
|
| 1814 |
+
"loss": 1.262,
|
| 1815 |
+
"step": 2390
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.384615384615385,
|
| 1819 |
+
"grad_norm": 1.1086236238479614,
|
| 1820 |
+
"learning_rate": 3.094545454545455e-05,
|
| 1821 |
+
"loss": 1.2275,
|
| 1822 |
+
"step": 2400
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.384615384615385,
|
| 1826 |
+
"eval_gen_len": 187.0146,
|
| 1827 |
+
"eval_loss": 1.6998823881149292,
|
| 1828 |
+
"eval_rouge1": 0.6055,
|
| 1829 |
+
"eval_rouge2": 0.2541,
|
| 1830 |
+
"eval_rougeL": 0.4684,
|
| 1831 |
+
"eval_runtime": 33.5145,
|
| 1832 |
+
"eval_samples_per_second": 8.176,
|
| 1833 |
+
"eval_steps_per_second": 2.059,
|
| 1834 |
+
"step": 2400
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 7.415384615384616,
|
| 1838 |
+
"grad_norm": 1.5682780742645264,
|
| 1839 |
+
"learning_rate": 3.0581818181818185e-05,
|
| 1840 |
+
"loss": 1.3442,
|
| 1841 |
+
"step": 2410
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 7.446153846153846,
|
| 1845 |
+
"grad_norm": 1.034818410873413,
|
| 1846 |
+
"learning_rate": 3.021818181818182e-05,
|
| 1847 |
+
"loss": 1.2158,
|
| 1848 |
+
"step": 2420
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 7.476923076923077,
|
| 1852 |
+
"grad_norm": 1.2816352844238281,
|
| 1853 |
+
"learning_rate": 2.985454545454546e-05,
|
| 1854 |
+
"loss": 1.1646,
|
| 1855 |
+
"step": 2430
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 7.507692307692308,
|
| 1859 |
+
"grad_norm": 1.252765417098999,
|
| 1860 |
+
"learning_rate": 2.9490909090909092e-05,
|
| 1861 |
+
"loss": 1.1985,
|
| 1862 |
+
"step": 2440
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 7.538461538461538,
|
| 1866 |
+
"grad_norm": 1.4074809551239014,
|
| 1867 |
+
"learning_rate": 2.9127272727272732e-05,
|
| 1868 |
+
"loss": 1.3245,
|
| 1869 |
+
"step": 2450
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 7.569230769230769,
|
| 1873 |
+
"grad_norm": 1.3757801055908203,
|
| 1874 |
+
"learning_rate": 2.8763636363636366e-05,
|
| 1875 |
+
"loss": 1.2856,
|
| 1876 |
+
"step": 2460
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 7.6,
|
| 1880 |
+
"grad_norm": 1.355635643005371,
|
| 1881 |
+
"learning_rate": 2.84e-05,
|
| 1882 |
+
"loss": 1.229,
|
| 1883 |
+
"step": 2470
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 7.63076923076923,
|
| 1887 |
+
"grad_norm": 1.185659646987915,
|
| 1888 |
+
"learning_rate": 2.803636363636364e-05,
|
| 1889 |
+
"loss": 1.2444,
|
| 1890 |
+
"step": 2480
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 7.661538461538462,
|
| 1894 |
+
"grad_norm": 1.4726060628890991,
|
| 1895 |
+
"learning_rate": 2.7672727272727273e-05,
|
| 1896 |
+
"loss": 1.2877,
|
| 1897 |
+
"step": 2490
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 7.6923076923076925,
|
| 1901 |
+
"grad_norm": 1.525718092918396,
|
| 1902 |
+
"learning_rate": 2.7309090909090913e-05,
|
| 1903 |
+
"loss": 1.2993,
|
| 1904 |
+
"step": 2500
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 7.723076923076923,
|
| 1908 |
+
"grad_norm": 1.2432451248168945,
|
| 1909 |
+
"learning_rate": 2.6945454545454546e-05,
|
| 1910 |
+
"loss": 1.2803,
|
| 1911 |
+
"step": 2510
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 7.753846153846154,
|
| 1915 |
+
"grad_norm": 1.2237839698791504,
|
| 1916 |
+
"learning_rate": 2.6581818181818186e-05,
|
| 1917 |
+
"loss": 1.2785,
|
| 1918 |
+
"step": 2520
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 7.7846153846153845,
|
| 1922 |
+
"grad_norm": 1.5360924005508423,
|
| 1923 |
+
"learning_rate": 2.621818181818182e-05,
|
| 1924 |
+
"loss": 1.3164,
|
| 1925 |
+
"step": 2530
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 7.815384615384615,
|
| 1929 |
+
"grad_norm": 1.4242442846298218,
|
| 1930 |
+
"learning_rate": 2.5854545454545453e-05,
|
| 1931 |
+
"loss": 1.1864,
|
| 1932 |
+
"step": 2540
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 7.846153846153846,
|
| 1936 |
+
"grad_norm": 1.0664770603179932,
|
| 1937 |
+
"learning_rate": 2.5490909090909093e-05,
|
| 1938 |
+
"loss": 1.2769,
|
| 1939 |
+
"step": 2550
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 7.876923076923077,
|
| 1943 |
+
"grad_norm": 1.3427962064743042,
|
| 1944 |
+
"learning_rate": 2.5127272727272727e-05,
|
| 1945 |
+
"loss": 1.183,
|
| 1946 |
+
"step": 2560
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 7.907692307692308,
|
| 1950 |
+
"grad_norm": 1.2692787647247314,
|
| 1951 |
+
"learning_rate": 2.4763636363636363e-05,
|
| 1952 |
+
"loss": 1.3775,
|
| 1953 |
+
"step": 2570
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 7.938461538461539,
|
| 1957 |
+
"grad_norm": 1.3220490217208862,
|
| 1958 |
+
"learning_rate": 2.44e-05,
|
| 1959 |
+
"loss": 1.1557,
|
| 1960 |
+
"step": 2580
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 7.969230769230769,
|
| 1964 |
+
"grad_norm": 1.31517493724823,
|
| 1965 |
+
"learning_rate": 2.4036363636363637e-05,
|
| 1966 |
+
"loss": 1.3144,
|
| 1967 |
+
"step": 2590
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 8.0,
|
| 1971 |
+
"grad_norm": 1.145848274230957,
|
| 1972 |
+
"learning_rate": 2.3672727272727274e-05,
|
| 1973 |
+
"loss": 1.4194,
|
| 1974 |
+
"step": 2600
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 8.0,
|
| 1978 |
+
"eval_gen_len": 187.0146,
|
| 1979 |
+
"eval_loss": 1.6980103254318237,
|
| 1980 |
+
"eval_rouge1": 0.6066,
|
| 1981 |
+
"eval_rouge2": 0.2565,
|
| 1982 |
+
"eval_rougeL": 0.4699,
|
| 1983 |
+
"eval_runtime": 33.5275,
|
| 1984 |
+
"eval_samples_per_second": 8.172,
|
| 1985 |
+
"eval_steps_per_second": 2.058,
|
| 1986 |
+
"step": 2600
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 8.03076923076923,
|
| 1990 |
+
"grad_norm": 1.3528770208358765,
|
| 1991 |
+
"learning_rate": 2.330909090909091e-05,
|
| 1992 |
+
"loss": 1.2504,
|
| 1993 |
+
"step": 2610
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 8.061538461538461,
|
| 1997 |
+
"grad_norm": 1.1651642322540283,
|
| 1998 |
+
"learning_rate": 2.2945454545454547e-05,
|
| 1999 |
+
"loss": 1.0993,
|
| 2000 |
+
"step": 2620
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 8.092307692307692,
|
| 2004 |
+
"grad_norm": 1.1845202445983887,
|
| 2005 |
+
"learning_rate": 2.258181818181818e-05,
|
| 2006 |
+
"loss": 1.1356,
|
| 2007 |
+
"step": 2630
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 8.123076923076923,
|
| 2011 |
+
"grad_norm": 1.2000699043273926,
|
| 2012 |
+
"learning_rate": 2.2218181818181817e-05,
|
| 2013 |
+
"loss": 1.206,
|
| 2014 |
+
"step": 2640
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 8.153846153846153,
|
| 2018 |
+
"grad_norm": 1.449044108390808,
|
| 2019 |
+
"learning_rate": 2.1854545454545454e-05,
|
| 2020 |
+
"loss": 1.2059,
|
| 2021 |
+
"step": 2650
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 8.184615384615384,
|
| 2025 |
+
"grad_norm": 1.2176152467727661,
|
| 2026 |
+
"learning_rate": 2.149090909090909e-05,
|
| 2027 |
+
"loss": 1.1849,
|
| 2028 |
+
"step": 2660
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 8.215384615384615,
|
| 2032 |
+
"grad_norm": 1.4765113592147827,
|
| 2033 |
+
"learning_rate": 2.1127272727272728e-05,
|
| 2034 |
+
"loss": 1.3351,
|
| 2035 |
+
"step": 2670
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 8.246153846153845,
|
| 2039 |
+
"grad_norm": 1.5038341283798218,
|
| 2040 |
+
"learning_rate": 2.0763636363636364e-05,
|
| 2041 |
+
"loss": 1.2766,
|
| 2042 |
+
"step": 2680
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 8.276923076923078,
|
| 2046 |
+
"grad_norm": 1.3483731746673584,
|
| 2047 |
+
"learning_rate": 2.04e-05,
|
| 2048 |
+
"loss": 1.1067,
|
| 2049 |
+
"step": 2690
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 8.307692307692308,
|
| 2053 |
+
"grad_norm": 1.025032639503479,
|
| 2054 |
+
"learning_rate": 2.0036363636363638e-05,
|
| 2055 |
+
"loss": 1.2155,
|
| 2056 |
+
"step": 2700
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 8.338461538461539,
|
| 2060 |
+
"grad_norm": 1.3824971914291382,
|
| 2061 |
+
"learning_rate": 1.9672727272727275e-05,
|
| 2062 |
+
"loss": 1.1767,
|
| 2063 |
+
"step": 2710
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 8.36923076923077,
|
| 2067 |
+
"grad_norm": 1.2280523777008057,
|
| 2068 |
+
"learning_rate": 1.930909090909091e-05,
|
| 2069 |
+
"loss": 1.2881,
|
| 2070 |
+
"step": 2720
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 8.4,
|
| 2074 |
+
"grad_norm": 1.2223644256591797,
|
| 2075 |
+
"learning_rate": 1.8945454545454548e-05,
|
| 2076 |
+
"loss": 1.1898,
|
| 2077 |
+
"step": 2730
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 8.430769230769231,
|
| 2081 |
+
"grad_norm": 1.349334955215454,
|
| 2082 |
+
"learning_rate": 1.8581818181818185e-05,
|
| 2083 |
+
"loss": 1.2984,
|
| 2084 |
+
"step": 2740
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 8.461538461538462,
|
| 2088 |
+
"grad_norm": 1.2894556522369385,
|
| 2089 |
+
"learning_rate": 1.8218181818181822e-05,
|
| 2090 |
+
"loss": 1.3458,
|
| 2091 |
+
"step": 2750
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 8.492307692307692,
|
| 2095 |
+
"grad_norm": 1.6086795330047607,
|
| 2096 |
+
"learning_rate": 1.7854545454545455e-05,
|
| 2097 |
+
"loss": 1.2394,
|
| 2098 |
+
"step": 2760
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 8.523076923076923,
|
| 2102 |
+
"grad_norm": 1.4955778121948242,
|
| 2103 |
+
"learning_rate": 1.7490909090909092e-05,
|
| 2104 |
+
"loss": 1.1506,
|
| 2105 |
+
"step": 2770
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 8.553846153846154,
|
| 2109 |
+
"grad_norm": 1.4156781435012817,
|
| 2110 |
+
"learning_rate": 1.712727272727273e-05,
|
| 2111 |
+
"loss": 1.1746,
|
| 2112 |
+
"step": 2780
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 8.584615384615384,
|
| 2116 |
+
"grad_norm": 1.2073826789855957,
|
| 2117 |
+
"learning_rate": 1.6763636363636365e-05,
|
| 2118 |
+
"loss": 1.1301,
|
| 2119 |
+
"step": 2790
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 8.615384615384615,
|
| 2123 |
+
"grad_norm": 1.574342966079712,
|
| 2124 |
+
"learning_rate": 1.6400000000000002e-05,
|
| 2125 |
+
"loss": 1.3976,
|
| 2126 |
+
"step": 2800
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 8.615384615384615,
|
| 2130 |
+
"eval_gen_len": 187.0146,
|
| 2131 |
+
"eval_loss": 1.7040081024169922,
|
| 2132 |
+
"eval_rouge1": 0.6049,
|
| 2133 |
+
"eval_rouge2": 0.2551,
|
| 2134 |
+
"eval_rougeL": 0.469,
|
| 2135 |
+
"eval_runtime": 34.1499,
|
| 2136 |
+
"eval_samples_per_second": 8.023,
|
| 2137 |
+
"eval_steps_per_second": 2.021,
|
| 2138 |
+
"step": 2800
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 8.646153846153846,
|
| 2142 |
+
"grad_norm": 1.207323670387268,
|
| 2143 |
+
"learning_rate": 1.603636363636364e-05,
|
| 2144 |
+
"loss": 1.2123,
|
| 2145 |
+
"step": 2810
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 8.676923076923076,
|
| 2149 |
+
"grad_norm": 1.346170425415039,
|
| 2150 |
+
"learning_rate": 1.5672727272727272e-05,
|
| 2151 |
+
"loss": 1.1284,
|
| 2152 |
+
"step": 2820
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 8.707692307692307,
|
| 2156 |
+
"grad_norm": 1.3920952081680298,
|
| 2157 |
+
"learning_rate": 1.530909090909091e-05,
|
| 2158 |
+
"loss": 1.1961,
|
| 2159 |
+
"step": 2830
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 8.73846153846154,
|
| 2163 |
+
"grad_norm": 1.4912207126617432,
|
| 2164 |
+
"learning_rate": 1.4945454545454546e-05,
|
| 2165 |
+
"loss": 1.2558,
|
| 2166 |
+
"step": 2840
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 8.76923076923077,
|
| 2170 |
+
"grad_norm": 0.9997207522392273,
|
| 2171 |
+
"learning_rate": 1.4581818181818183e-05,
|
| 2172 |
+
"loss": 1.1842,
|
| 2173 |
+
"step": 2850
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 8.8,
|
| 2177 |
+
"grad_norm": 1.207138180732727,
|
| 2178 |
+
"learning_rate": 1.421818181818182e-05,
|
| 2179 |
+
"loss": 1.2588,
|
| 2180 |
+
"step": 2860
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 8.830769230769231,
|
| 2184 |
+
"grad_norm": 1.398917555809021,
|
| 2185 |
+
"learning_rate": 1.3854545454545456e-05,
|
| 2186 |
+
"loss": 1.2445,
|
| 2187 |
+
"step": 2870
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 8.861538461538462,
|
| 2191 |
+
"grad_norm": 1.3793071508407593,
|
| 2192 |
+
"learning_rate": 1.3490909090909093e-05,
|
| 2193 |
+
"loss": 1.164,
|
| 2194 |
+
"step": 2880
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 8.892307692307693,
|
| 2198 |
+
"grad_norm": 1.2650920152664185,
|
| 2199 |
+
"learning_rate": 1.3127272727272726e-05,
|
| 2200 |
+
"loss": 1.2235,
|
| 2201 |
+
"step": 2890
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 8.923076923076923,
|
| 2205 |
+
"grad_norm": 1.3319740295410156,
|
| 2206 |
+
"learning_rate": 1.2763636363636363e-05,
|
| 2207 |
+
"loss": 1.1818,
|
| 2208 |
+
"step": 2900
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 8.953846153846154,
|
| 2212 |
+
"grad_norm": 1.395668387413025,
|
| 2213 |
+
"learning_rate": 1.24e-05,
|
| 2214 |
+
"loss": 1.2237,
|
| 2215 |
+
"step": 2910
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 8.984615384615385,
|
| 2219 |
+
"grad_norm": 1.1730422973632812,
|
| 2220 |
+
"learning_rate": 1.2036363636363637e-05,
|
| 2221 |
+
"loss": 1.0141,
|
| 2222 |
+
"step": 2920
|
| 2223 |
+
}
|
| 2224 |
+
],
|
| 2225 |
+
"logging_steps": 10,
|
| 2226 |
+
"max_steps": 3250,
|
| 2227 |
+
"num_input_tokens_seen": 0,
|
| 2228 |
+
"num_train_epochs": 10,
|
| 2229 |
+
"save_steps": 500,
|
| 2230 |
+
"stateful_callbacks": {
|
| 2231 |
+
"TrainerControl": {
|
| 2232 |
+
"args": {
|
| 2233 |
+
"should_epoch_stop": false,
|
| 2234 |
+
"should_evaluate": false,
|
| 2235 |
+
"should_log": false,
|
| 2236 |
+
"should_save": true,
|
| 2237 |
+
"should_training_stop": false
|
| 2238 |
+
},
|
| 2239 |
+
"attributes": {}
|
| 2240 |
+
}
|
| 2241 |
+
},
|
| 2242 |
+
"total_flos": 7124807319552000.0,
|
| 2243 |
+
"train_batch_size": 4,
|
| 2244 |
+
"trial_name": null,
|
| 2245 |
+
"trial_params": null
|
| 2246 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8dff354338ed1884db2b62929f25f690c2b444c83fcc581cc89e6c23bc8d7f2f
|
| 3 |
+
size 5240
|