Update PPO model

README.md

@@ -25,7 +25,7 @@ You can then generate text as follows:
 ```python
 from transformers import pipeline
 
-generator = pipeline("text-generation", model="KB8407//content/drive/MyDrive/R&E 2B (Good)/outputs/ppo/final")
+generator = pipeline("text-generation", model="KB8407/outputs/ppo/final")
 outputs = generator("Hello, my llama is cute")
 ```
 
@@ -35,8 +35,8 @@ If you want to use the model for training or to obtain the outputs from the valu
 from transformers import AutoTokenizer
 from trl import AutoModelForCausalLMWithValueHead
 
-tokenizer = AutoTokenizer.from_pretrained("KB8407//content/drive/MyDrive/R&E 2B (Good)/outputs/ppo/final")
-model = AutoModelForCausalLMWithValueHead.from_pretrained("KB8407//content/drive/MyDrive/R&E 2B (Good)/outputs/ppo/final")
+tokenizer = AutoTokenizer.from_pretrained("KB8407/outputs/ppo/final")
+model = AutoModelForCausalLMWithValueHead.from_pretrained("KB8407/outputs/ppo/final")
 
 inputs = tokenizer("Hello, my llama is cute", return_tensors="pt")
 outputs = model(**inputs, labels=inputs["input_ids"])

config.json

@@ -8,9 +8,9 @@
   "author": "Heewon Jeon(madjakarta@gmail.com)",
   "bos_token_id": 0,
   "created_date": "2021-04-28",
-  "dtype": "bfloat16",
+  "dtype": "float32",
   "embd_pdrop": 0.1,
-  "eos_token_id": 1,
+  "eos_token_id": 51200,
   "gradient_checkpointing": false,
   "id2label": {
     "0": "LABEL_0"
@@ -28,7 +28,7 @@
   "n_inner": null,
   "n_layer": 12,
   "n_positions": 1024,
-  "pad_token_id": 3,
+  "pad_token_id": 51200,
   "reorder_and_upcast_attn": false,
   "resid_pdrop": 0.1,
   "scale_attn_by_inverse_layer_idx": false,
@@ -44,7 +44,7 @@
       "max_length": 50
     }
   },
-  "transformers_version": "4.57.2",
+  "transformers_version": "4.57.3",
   "use_cache": true,
-  "vocab_size": 51200
+  "vocab_size": 51201
 }

generation_config.json

@@ -3,5 +3,5 @@
   "bos_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 3,
-  "transformers_version": "4.57.2"
+  "transformers_version": "4.57.3"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a584adb5bf54418da76a7feabe963a70d1cb7fb1ede012061c8aef95e6de738
-size 250346332
+oid sha256:7325c79dc2a434e6597b2df1e9c258546030f6c8f7a1bad45f372d8c31d2d564
+size 500677436

tokenizer_config.json

@@ -3014,8 +3014,12 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "<|endoftext|>",
   "extra_special_tokens": {},
+  "max_length": 512,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<|endoftext|>",
+  "stride": 0,
   "tokenizer_class": "GPT2Tokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
   "unk_token": "<|endoftext|>"
 }