Upload 6 files

README.md

@@ -0,0 +1,43 @@
+---
+license: apache-2.0
+tags:
+- trl
+- ppo
+- transformers
+- reinforcement-learning
+---
+
+# TRL Model
+
+This is a [TRL language model](https://github.com/huggingface/trl) that has been fine-tuned with reinforcement learning to
+ guide the model outputs according to a value, function, or human feedback. The model can be used for text generation.
+
+## Usage
+
+To use this model for inference, first install the TRL library:
+
+```bash
+python -m pip install trl
+```
+
+You can then generate text as follows:
+
+```python
+from transformers import pipeline
+
+generator = pipeline("text-generation", model="anirvankrishna/PPO/prefix_tuning/best_model")
+outputs = generator("Hello, my llama is cute")
+```
+
+If you want to use the model for training or to obtain the outputs from the value head, load the model as follows:
+
+```python
+from transformers import AutoTokenizer
+from trl import AutoModelForCausalLMWithValueHead
+
+tokenizer = AutoTokenizer.from_pretrained("anirvankrishna/PPO/prefix_tuning/best_model")
+model = AutoModelForCausalLMWithValueHead.from_pretrained("anirvankrishna/PPO/prefix_tuning/best_model")
+
+inputs = tokenizer("Hello, my llama is cute", return_tensors="pt")
+outputs = model(**inputs, labels=inputs["input_ids"])
+```

adapter_config.json

@@ -0,0 +1,17 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "openai-community/gpt2-medium",
+  "encoder_hidden_size": 1024,
+  "inference_mode": true,
+  "modules_to_save": null,
+  "num_attention_heads": 16,
+  "num_layers": 24,
+  "num_transformer_submodules": 1,
+  "num_virtual_tokens": 10,
+  "peft_type": "PREFIX_TUNING",
+  "peft_version": "0.18.1",
+  "prefix_projection": false,
+  "revision": null,
+  "task_type": "CAUSAL_LM",
+  "token_dim": 1024
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:016cca9f07cd111480fa8e706c0fc130f54028c519dbd4d74864a6e27c1e9ba7
+size 1966208

config.json

@@ -0,0 +1,49 @@
+{
+  "accelerator_kwargs": {},
+  "adap_kl_ctrl": true,
+  "backward_batch_size": 32,
+  "batch_size": 32,
+  "cliprange": 0.2,
+  "cliprange_value": 0.2,
+  "compare_steps": 1,
+  "early_stopping": false,
+  "exp_name": "colab_kernel_launcher",
+  "forward_batch_size": null,
+  "gamma": 1,
+  "global_backward_batch_size": 32,
+  "global_batch_size": 32,
+  "gradient_accumulation_steps": 1,
+  "horizon": 10000,
+  "init_kl_coef": 0.1,
+  "is_encoder_decoder": false,
+  "is_peft_model": true,
+  "kl_penalty": "kl",
+  "lam": 0.95,
+  "learning_rate": 0.0001,
+  "log_with": null,
+  "max_grad_norm": null,
+  "mini_batch_size": 32,
+  "model_name": "gpt2",
+  "optimize_cuda_cache": true,
+  "optimize_device_cache": false,
+  "ppo_epochs": 1,
+  "project_kwargs": {},
+  "push_to_hub_if_best_kwargs": {},
+  "query_dataset": "imdb",
+  "ratio_threshold": 10.0,
+  "remove_unused_columns": true,
+  "reward_model": "sentiment-analysis:lvwerra/distilbert-imdb",
+  "score_clip": null,
+  "seed": 0,
+  "steps": 10,
+  "target": 0.1,
+  "target_kl": 1,
+  "task_name": null,
+  "tracker_kwargs": {},
+  "tracker_project_name": "trl",
+  "use_score_norm": false,
+  "use_score_scaling": false,
+  "vf_coef": 0.1,
+  "whiten_rewards": false,
+  "world_size": 1
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d296d8763f5ef2e38958c1ec2228c39025c4700fcf5c1a879da051c7a4844ce
+size 6037