Training in progress, step 1000

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "facebook/opt-125m",
+  "_name_or_path": "facebook/opt-350m",
   "_remove_final_layer_norm": false,
   "activation_dropout": 0.0,
   "activation_function": "relu",
@@ -8,12 +8,12 @@
   ],
   "attention_dropout": 0.0,
   "bos_token_id": 2,
-  "do_layer_norm_before": true,
+  "do_layer_norm_before": false,
   "dropout": 0.1,
   "enable_bias": true,
   "eos_token_id": 2,
-  "ffn_dim": 3072,
-  "hidden_size": 768,
+  "ffn_dim": 4096,
+  "hidden_size": 1024,
   "id2label": {
     "0": "negative",
     "1": "positive"
@@ -27,8 +27,8 @@
   "layerdrop": 0.0,
   "max_position_embeddings": 2048,
   "model_type": "opt",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
   "pad_token_id": 1,
   "prefix": "</s>",
   "problem_type": "single_label_classification",
@@ -36,5 +36,5 @@
   "transformers_version": "4.43.3",
   "use_cache": true,
   "vocab_size": 50272,
-  "word_embed_proj_dim": 768
+  "word_embed_proj_dim": 512
 }

logs/events.out.tfevents.1722352771.163db1aa8eaf.2333.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec0cd659c501310eb6145deba490bec9fe4686d18f79c36a58500be8cd4e2474
-size 5713
+oid sha256:36c74985f8363ae2df4083b626c6f54267ba76751f12d3b7f98e254252e06066
+size 6067

logs/events.out.tfevents.1722353866.163db1aa8eaf.7613.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0378039692f01f95189f031f4bd171a41959ac4aff9670fcb69a63731086f80
+size 4184

logs/events.out.tfevents.1722353959.163db1aa8eaf.8429.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d4f0ef1e38c47974327fe9aa13b01c63ccb4995399109877963e24eb44f63fe
+size 5132

logs/events.out.tfevents.1722353972.163db1aa8eaf.8429.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6f3619bd7d1f2d6b6bcc5dbfb92ede94c2e4f971e39a3b0f08f71134c9911d8
+size 5132

logs/events.out.tfevents.1722353982.163db1aa8eaf.8429.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aaf3195f37001062e8c88497b0fe6c065d76e9b5d94a9ad35fe3df8d7eb437d
+size 5666

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1089ad5ca315c54f5915342212832d885e5227c028120fcee0d71fce2f2c5eaf
-size 500985832
+oid sha256:839a39b7843b7cb1d62e9a2350a852a9a2679b815f8be5e778c47a42ec7281a8
+size 1324835064

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:077a20e9d853074663a32801a4612ac716bb32a6a8a9d14e943bf4bfc451ab96
+oid sha256:c03ccc6f5f6a85743f13c9b224b035feaaaa9b018ca64cf2c384d70760b6811c
 size 5240