Training in progress, step 20
Browse files- config.json +46 -0
- model.safetensors +3 -0
- runs/Aug23_00-32-05_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724353325.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.112323.0 +3 -0
- runs/Aug23_15-04-56_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405697.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116698.0 +3 -0
- runs/Aug23_15-06-53_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405814.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116769.0 +3 -0
- runs/Aug23_15-08-23_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405904.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116830.0 +3 -0
- runs/Aug23_15-09-37_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405977.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116889.0 +3 -0
- runs/Aug23_15-11-41_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724406102.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116967.0 +3 -0
- training_args.bin +3 -0
config.json
ADDED
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "ai4bharat/indictrans2-en-indic-dist-200M",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "gelu",
|
5 |
+
"architectures": [
|
6 |
+
"IndicTransForConditionalGeneration"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.0,
|
9 |
+
"attn_implementation": null,
|
10 |
+
"auto_map": {
|
11 |
+
"AutoConfig": "ai4bharat/indictrans2-en-indic-dist-200M--configuration_indictrans.IndicTransConfig",
|
12 |
+
"AutoModelForSeq2SeqLM": "ai4bharat/indictrans2-en-indic-dist-200M--modeling_indictrans.IndicTransForConditionalGeneration"
|
13 |
+
},
|
14 |
+
"bos_token_id": 0,
|
15 |
+
"decoder_attention_heads": 8,
|
16 |
+
"decoder_embed_dim": 512,
|
17 |
+
"decoder_ffn_dim": 2048,
|
18 |
+
"decoder_layerdrop": 0,
|
19 |
+
"decoder_layers": 18,
|
20 |
+
"decoder_normalize_before": true,
|
21 |
+
"decoder_start_token_id": 2,
|
22 |
+
"decoder_vocab_size": 122672,
|
23 |
+
"dropout": 0.2,
|
24 |
+
"encoder_attention_heads": 8,
|
25 |
+
"encoder_embed_dim": 512,
|
26 |
+
"encoder_ffn_dim": 2048,
|
27 |
+
"encoder_layerdrop": 0,
|
28 |
+
"encoder_layers": 18,
|
29 |
+
"encoder_normalize_before": true,
|
30 |
+
"encoder_vocab_size": 32322,
|
31 |
+
"eos_token_id": 2,
|
32 |
+
"init_std": 0.02,
|
33 |
+
"is_encoder_decoder": true,
|
34 |
+
"layernorm_embedding": true,
|
35 |
+
"max_source_positions": 256,
|
36 |
+
"max_target_positions": 256,
|
37 |
+
"model_type": "IndicTrans",
|
38 |
+
"num_hidden_layers": 18,
|
39 |
+
"pad_token_id": 1,
|
40 |
+
"scale_embedding": true,
|
41 |
+
"share_decoder_input_output_embed": true,
|
42 |
+
"tokenizer_class": "IndicTransTokenizer",
|
43 |
+
"torch_dtype": "bfloat16",
|
44 |
+
"transformers_version": "4.44.0",
|
45 |
+
"use_cache": true
|
46 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f737aa413837de162bcd5c24973bba6f116a1736dcc653dea3ff74eb0a5b887e
|
3 |
+
size 549258952
|
runs/Aug23_00-32-05_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724353325.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.112323.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:86b95db0fd32b4ae8e41eaeebe8fb54d5537346a81ad4c971d3e96960d0cfb07
|
3 |
+
size 48103
|
runs/Aug23_15-04-56_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405697.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116698.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9614ba8a9045b291595d451131fc7dd689f3c844ea39a25f80987efcdf49c16f
|
3 |
+
size 5919
|
runs/Aug23_15-06-53_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405814.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116769.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:42a584fc57f15a7ab6b88645a16fe783a36312772f287ac1b174a29b1a7df2fa
|
3 |
+
size 5920
|
runs/Aug23_15-08-23_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405904.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116830.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d6fe5a11cedc3b54ea0d4041b5ecd6e088180280f91095399268ec28d8627215
|
3 |
+
size 5921
|
runs/Aug23_15-09-37_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724405977.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116889.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:67bdfe8583aebed7aabd45f55e9453e02b99fb5261e9467d0d50b36e7a2c647b
|
3 |
+
size 5921
|
runs/Aug23_15-11-41_machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC/events.out.tfevents.1724406102.machine1-HP-Z2-Tower-G9-Workstation-Desktop-PC.116967.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d6a3fcaf75cb091c242e56e397f98eebd86c3769a45abc917bf79b2f0898220
|
3 |
+
size 6452
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f1b9de20ea5e9bcbac44cfdc5a7eceab986294acf3257afd6b2a9ec119750b33
|
3 |
+
size 5368
|