Push model using huggingface_hub.

README.md

@@ -2,6 +2,7 @@
 license: apache-2.0
 tags:
 - trl
+- ppo
 - transformers
 - reinforcement-learning
 ---
@@ -24,7 +25,7 @@ You can then generate text as follows:
 ```python
 from transformers import pipeline
 
-generator = pipeline("text-generation", model="damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmpi3wg7drx/damienbenveniste/mistral-ppo")
+generator = pipeline("text-generation", model="damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmppue5zznd/damienbenveniste/mistral-ppo")
 outputs = generator("Hello, my llama is cute")
 ```
 
@@ -34,8 +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("damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmpi3wg7drx/damienbenveniste/mistral-ppo")
-model = AutoModelForCausalLMWithValueHead.from_pretrained("damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmpi3wg7drx/damienbenveniste/mistral-ppo")
+tokenizer = AutoTokenizer.from_pretrained("damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmppue5zznd/damienbenveniste/mistral-ppo")
+model = AutoModelForCausalLMWithValueHead.from_pretrained("damienbenveniste//var/folders/qj/lfvfq6590q5fn7hnwx6c29k80000gn/T/tmppue5zznd/damienbenveniste/mistral-ppo")
 
 inputs = tokenizer("Hello, my llama is cute", return_tensors="pt")
 outputs = model(**inputs, labels=inputs["input_ids"])

config.json

@@ -9,14 +9,11 @@
   "hidden_size": 768,
   "initializer_range": 0.02,
   "intermediate_size": 3072,
-  "max_length": 512,
   "max_position_embeddings": 512,
   "model_type": "mistral",
   "num_attention_heads": 16,
   "num_hidden_layers": 4,
   "num_key_value_heads": 8,
-  "pad_token_id": 32000,
-  "padding_side": "left",
   "rms_norm_eps": 1e-06,
   "rope_theta": 10000.0,
   "sliding_window": 768,

generation_config.json

@@ -2,7 +2,5 @@
   "_from_model_config": true,
   "bos_token_id": 1,
   "eos_token_id": 2,
-  "max_length": 512,
-  "pad_token_id": 32000,
   "transformers_version": "4.35.2"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:883cee707da1249523d4b9366967a0f215c5ee2ee3bd377cb5425f975e5a02c3
+oid sha256:d8de2bb878800087d0c7fbc318e19ee4fc1d428a587daa2bab14ccc6363b42ae
 size 338200972

special_tokens_map.json

@@ -14,7 +14,7 @@
     "single_word": false
   },
   "pad_token": {
-    "content": "<pad>",
+    "content": "</s>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,

tokenizer.json

@@ -1,21 +1,7 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 512,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
-  "padding": {
-    "strategy": {
-      "Fixed": 512
-    },
-    "direction": "Left",
-    "pad_to_multiple_of": null,
-    "pad_id": 32000,
-    "pad_type_id": 0,
-    "pad_token": "<pad>"
-  },
+  "truncation": null,
+  "padding": null,
   "added_tokens": [
     {
       "id": 0,
@@ -43,15 +29,6 @@
       "rstrip": false,
       "normalized": false,
       "special": true
-    },
-    {
-      "id": 32000,
-      "content": "<pad>",
-      "single_word": false,
-      "lstrip": false,
-      "rstrip": false,
-      "normalized": false,
-      "special": true
     }
   ],
   "normalizer": {

tokenizer_config.json

@@ -23,14 +23,6 @@
       "rstrip": false,
       "single_word": false,
       "special": true
-    },
-    "32000": {
-      "content": "<pad>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
     }
   },
   "additional_special_tokens": [],
@@ -41,9 +33,8 @@
   "max_length": 30000,
   "model_max_length": 1000000000000000019884624838656,
   "pad_to_multiple_of": null,
-  "pad_token": "<pad>",
+  "pad_token": "</s>",
   "pad_token_type_id": 0,
-  "paddind_side": "left",
   "padding_side": "left",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,