Training in progress, step 500
Browse files- .gitattributes +1 -0
- .gitignore +1 -0
- config.json +31 -0
- pytorch_model.bin +3 -0
- runs/May19_14-02-53_9b371da60e9a/1684504978.8468924/events.out.tfevents.1684504978.9b371da60e9a.163.3 +3 -0
- runs/May19_14-02-53_9b371da60e9a/events.out.tfevents.1684504978.9b371da60e9a.163.2 +3 -0
- runs/May19_14-14-46_9b371da60e9a/1684505688.4917831/events.out.tfevents.1684505688.9b371da60e9a.163.5 +3 -0
- runs/May19_14-14-46_9b371da60e9a/events.out.tfevents.1684505688.9b371da60e9a.163.4 +3 -0
- runs/May19_14-17-18_9b371da60e9a/1684505844.40966/events.out.tfevents.1684505844.9b371da60e9a.7004.1 +3 -0
- runs/May19_14-17-18_9b371da60e9a/events.out.tfevents.1684505844.9b371da60e9a.7004.0 +3 -0
- runs/May19_14-19-29_9b371da60e9a/1684505974.898217/events.out.tfevents.1684505974.9b371da60e9a.7815.1 +3 -0
- runs/May19_14-19-29_9b371da60e9a/events.out.tfevents.1684505974.9b371da60e9a.7815.0 +3 -0
- runs/May19_14-21-15_9b371da60e9a/1684506079.515968/events.out.tfevents.1684506079.9b371da60e9a.8511.1 +3 -0
- runs/May19_14-21-15_9b371da60e9a/events.out.tfevents.1684506079.9b371da60e9a.8511.0 +3 -0
- special_tokens_map.json +5 -0
- spiece.model +3 -0
- tokenizer.json +3 -0
- tokenizer_config.json +11 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 35 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
.gitignore
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
checkpoint-*/
|
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/mt5-small",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MT5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"d_ff": 1024,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 512,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "gelu_new",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "gated-gelu",
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"is_encoder_decoder": true,
|
| 16 |
+
"is_gated_act": true,
|
| 17 |
+
"layer_norm_epsilon": 1e-06,
|
| 18 |
+
"model_type": "mt5",
|
| 19 |
+
"num_decoder_layers": 8,
|
| 20 |
+
"num_heads": 6,
|
| 21 |
+
"num_layers": 8,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"relative_attention_max_distance": 128,
|
| 24 |
+
"relative_attention_num_buckets": 32,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"tokenizer_class": "T5Tokenizer",
|
| 27 |
+
"torch_dtype": "float32",
|
| 28 |
+
"transformers_version": "4.28.0",
|
| 29 |
+
"use_cache": true,
|
| 30 |
+
"vocab_size": 250112
|
| 31 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47be59e75c69d4f3e2e0872513151dcb823dab759d169c384d6c2d3fec388ab7
|
| 3 |
+
size 1200772485
|
runs/May19_14-02-53_9b371da60e9a/1684504978.8468924/events.out.tfevents.1684504978.9b371da60e9a.163.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c952c7f99979230884a8159e253197b2df768228b55ffa0bbe08c4886f101430
|
| 3 |
+
size 6188
|
runs/May19_14-02-53_9b371da60e9a/events.out.tfevents.1684504978.9b371da60e9a.163.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56e24e4cfaa35d160b141400753893fb960458ea155f9f8eaa5db48536b32a3a
|
| 3 |
+
size 6719
|
runs/May19_14-14-46_9b371da60e9a/1684505688.4917831/events.out.tfevents.1684505688.9b371da60e9a.163.5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6fb05155c5e1bf4c6910b72df8378d800de3f00d77b078f8b296f90175d18ad
|
| 3 |
+
size 6184
|
runs/May19_14-14-46_9b371da60e9a/events.out.tfevents.1684505688.9b371da60e9a.163.4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:536365f2958017d15b24ab1eb3b4a51e508b35dbdc545953b416a41780635516
|
| 3 |
+
size 4184
|
runs/May19_14-17-18_9b371da60e9a/1684505844.40966/events.out.tfevents.1684505844.9b371da60e9a.7004.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d87acdb300bfd655f4c2bcbaf1140f7de4eb6f0402aa42150d8edfc3d5554bf
|
| 3 |
+
size 6184
|
runs/May19_14-17-18_9b371da60e9a/events.out.tfevents.1684505844.9b371da60e9a.7004.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e42c66e604e618d48db4d10b057146155b213ba593d0f259dfeff9984d9019a
|
| 3 |
+
size 4184
|
runs/May19_14-19-29_9b371da60e9a/1684505974.898217/events.out.tfevents.1684505974.9b371da60e9a.7815.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44bc1b7355135e851acae034219927ca62c6368ab98c49736e7ecc9e9b5e0baf
|
| 3 |
+
size 6184
|
runs/May19_14-19-29_9b371da60e9a/events.out.tfevents.1684505974.9b371da60e9a.7815.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c67cdc48d8067d5249cde9841080b6797ac5d59e6436b25abb01aeb69353e3a4
|
| 3 |
+
size 4184
|
runs/May19_14-21-15_9b371da60e9a/1684506079.515968/events.out.tfevents.1684506079.9b371da60e9a.8511.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24bf3e91043aa65bbb2c637afcbe6ce7919e3c2004894d55c1f1be5ecb9c3bb6
|
| 3 |
+
size 6184
|
runs/May19_14-21-15_9b371da60e9a/events.out.tfevents.1684506079.9b371da60e9a.8511.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8352ffb44a6d245e0cf7f7b489a1d80b2ca47def4c62b0b0ebdde3ae518db2b0
|
| 3 |
+
size 5479
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "</s>",
|
| 3 |
+
"pad_token": "<pad>",
|
| 4 |
+
"unk_token": "<unk>"
|
| 5 |
+
}
|
spiece.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
|
| 3 |
+
size 4309802
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:faaa6405f5f79c9e788c7980874a9a3b5b0aea07b53bd9243bf1abb8f5c49c81
|
| 3 |
+
size 16330467
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": null,
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"extra_ids": 0,
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"pad_token": "<pad>",
|
| 8 |
+
"sp_model_kwargs": {},
|
| 9 |
+
"tokenizer_class": "T5Tokenizer",
|
| 10 |
+
"unk_token": "<unk>"
|
| 11 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9884b4282a143d62d1e8499e15ee32d3286a7065e65a7ea4bac851b61c218dd5
|
| 3 |
+
size 3771
|