Training in progress, step 500

config.json

@@ -1,19 +1,19 @@
 {
-  "_name_or_path": "theojolliffe/T5-model-1-feedback-1109",
+  "_name_or_path": "google/flan-t5-base",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
-  "d_ff": 3072,
+  "d_ff": 2048,
   "d_kv": 64,
   "d_model": 768,
   "decoder_start_token_id": 0,
-  "dense_act_fn": "relu",
+  "dense_act_fn": "gelu_new",
   "dropout_rate": 0.1,
   "eos_token_id": 1,
-  "feed_forward_proj": "relu",
+  "feed_forward_proj": "gated-gelu",
   "initializer_factor": 1.0,
   "is_encoder_decoder": true,
-  "is_gated_act": false,
+  "is_gated_act": true,
   "layer_norm_epsilon": 1e-06,
   "model_type": "t5",
   "n_positions": 512,
@@ -53,8 +53,9 @@
       "prefix": "translate English to Romanian: "
     }
   },
+  "tie_word_embeddings": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.24.0",
+  "transformers_version": "4.26.1",
   "use_cache": true,
   "vocab_size": 32128
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2645ac0cb4cdaacc3ee6cbc5258fea22339ca296b8b8f6d895bbd2bf7b8db200
-size 891700799
+oid sha256:a9c41f877fe0e15973fabd60eebe3564b42af1d4ff106ce50b28fe6c667b6cb6
+size 990408885

runs/Feb28_23-14-29_5d14e5e5ecd4/1677626320.1970775/events.out.tfevents.1677626320.5d14e5e5ecd4.1155.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0b0f99a134a3d6d94d2be52b7b1f8493e6fc80630919f2255ddce69b8db5a34
+size 5933

runs/Feb28_23-14-29_5d14e5e5ecd4/events.out.tfevents.1677626320.5d14e5e5ecd4.1155.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76d290485ae87db696c1a0cd280885b1b42834f25662c789827c59ab6760b80c
+size 5505

tokenizer_config.json

@@ -104,11 +104,10 @@
   "eos_token": "</s>",
   "extra_ids": 100,
   "model_max_length": 512,
-  "name_or_path": "theojolliffe/T5-model-1-feedback-1109",
+  "name_or_path": "google/flan-t5-base",
   "pad_token": "<pad>",
   "sp_model_kwargs": {},
-  "special_tokens_map_file": null,
+  "special_tokens_map_file": "/home/younes_huggingface_co/.cache/huggingface/hub/models--google--t5-v1_1-base/snapshots/650d7745bf1e502d6949b22cc19155cd656d3d4e/special_tokens_map.json",
   "tokenizer_class": "T5Tokenizer",
-  "tokenizer_file": null,
   "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da20caf121ffcdb7830c09ebc1252503fcf682bbe39fb501404f77e6db654352
-size 3503
+oid sha256:13cbb44cbbba7056ee120e6ef4ee00e1a6988eb8d5b0fe49ae87c448277eca53
+size 3643