Upload 3 files
Browse files- all_results.json +8 -0
- config.json +55 -0
- generation_config.json +8 -0
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"total_flos": 6.099787643229307e+18,
|
4 |
+
"train_loss": 0.04448508482245479,
|
5 |
+
"train_runtime": 11326.9906,
|
6 |
+
"train_samples_per_second": 132.859,
|
7 |
+
"train_steps_per_second": 1.038
|
8 |
+
}
|
config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "/HOME/uestc_jksong/uestc_jksong_1/HDD_POOL/zhanghaonan/huggingface/paligemma2-3b-pt-224",
|
3 |
+
"_vocab_size": 257152,
|
4 |
+
"architectures": [
|
5 |
+
"PaliGemmaForConditionalGeneration"
|
6 |
+
],
|
7 |
+
"bos_token_id": 2,
|
8 |
+
"eos_token_id": 1,
|
9 |
+
"hidden_size": 2048,
|
10 |
+
"image_token_index": 257152,
|
11 |
+
"model_type": "paligemma",
|
12 |
+
"num_hidden_layers": 26,
|
13 |
+
"pad_token_id": 0,
|
14 |
+
"projection_dim": 2304,
|
15 |
+
"text_config": {
|
16 |
+
"architectures": [
|
17 |
+
"Gemma2ForCausalLM"
|
18 |
+
],
|
19 |
+
"attn_logit_softcapping": 50.0,
|
20 |
+
"cache_implementation": "hybrid",
|
21 |
+
"eos_token_id": [
|
22 |
+
1,
|
23 |
+
107
|
24 |
+
],
|
25 |
+
"final_logit_softcapping": 30.0,
|
26 |
+
"hidden_act": "gelu_pytorch_tanh",
|
27 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
28 |
+
"hidden_size": 2304,
|
29 |
+
"intermediate_size": 9216,
|
30 |
+
"model_type": "gemma2",
|
31 |
+
"num_hidden_layers": 26,
|
32 |
+
"num_image_tokens": 256,
|
33 |
+
"num_key_value_heads": 4,
|
34 |
+
"query_pre_attn_scalar": 256,
|
35 |
+
"sliding_window": 4096,
|
36 |
+
"torch_dtype": "bfloat16",
|
37 |
+
"vocab_size": 257216
|
38 |
+
},
|
39 |
+
"torch_dtype": "bfloat16",
|
40 |
+
"transformers_version": "4.49.0",
|
41 |
+
"use_cache": false,
|
42 |
+
"vision_config": {
|
43 |
+
"hidden_size": 1152,
|
44 |
+
"intermediate_size": 4304,
|
45 |
+
"model_type": "siglip_vision_model",
|
46 |
+
"num_attention_heads": 16,
|
47 |
+
"num_hidden_layers": 27,
|
48 |
+
"num_image_tokens": 256,
|
49 |
+
"num_positions": 256,
|
50 |
+
"patch_size": 14,
|
51 |
+
"projection_dim": 2304,
|
52 |
+
"torch_dtype": "bfloat16",
|
53 |
+
"vision_use_head": false
|
54 |
+
}
|
55 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 2,
|
4 |
+
"cache_implementation": "hybrid",
|
5 |
+
"eos_token_id": 1,
|
6 |
+
"pad_token_id": 0,
|
7 |
+
"transformers_version": "4.49.0"
|
8 |
+
}
|