feat: added actor model

README.md

@@ -13,7 +13,7 @@ license: apache-2.0
 
 # ChatGPT OPT 1.3B DeepSpeed Reinforcement Learning from Human feedback
 
-*fsalab-chat-opt-1.3b-rlhf-deepspeed*
+*fsalab-chat-opt-1.3b-rlhf-actor-deepspeed*
 
 This model consists of the final step of a modified pipeline the to the traditional training process of Chat-GPT models, which is comprised of a three-step procedure of [supervised fine tuning](https://huggingface.co/FSALab/fsalab-chat-opt-1.3b-sft-deepspeed), [reward model](https://huggingface.co/FSALab/fsalab-chat-opt-350m-reward-deepspeed) and **reinforcement learning from human feedback**.
 

config.json

@@ -0,0 +1,32 @@
+{
+  "_name_or_path": "/lus/grand/projects/BNN-Scale/chatgpt/hf_runs/DeepSpeedExamples/applications/DeepSpeed-Chat/output/actor-models/1.3b",
+  "_remove_final_layer_norm": false,
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "OPTForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 2,
+  "do_layer_norm_before": true,
+  "dropout": 0.0,
+  "enable_bias": true,
+  "end_token_id": 50260,
+  "eos_token_id": 2,
+  "ffn_dim": 8192,
+  "hidden_size": 2048,
+  "init_std": 0.02,
+  "layer_norm_elementwise_affine": true,
+  "layerdrop": 0.0,
+  "max_position_embeddings": 2048,
+  "model_type": "opt",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 24,
+  "pad_token_id": 2,
+  "prefix": "</s>",
+  "torch_dtype": "float16",
+  "transformers_version": "4.29.0.dev0",
+  "use_cache": true,
+  "vocab_size": 50272,
+  "word_embed_proj_dim": 2048
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cbb17038fa906783a3dee89e31c89c1cef03cea47723a5f065cc7b78ca37c89
+size 3235869199