Upload 7 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/qlora/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/qlora/best_model")
+model = AutoModelForCausalLMWithValueHead.from_pretrained("anirvankrishna/PPO/qlora/best_model")
+
+inputs = tokenizer("Hello, my llama is cute", return_tensors="pt")
+outputs = model(**inputs, labels=inputs["input_ids"])
+```

adapter_config.json

@@ -0,0 +1,40 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "openai-community/gpt2-medium",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": true,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 8,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff39f95de294c5a6cbd82323611317d1e0278e718beae46996bcda5a1f0379d6
+size 3151856

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": 0.5,
+  "is_encoder_decoder": false,
+  "is_peft_model": true,
+  "kl_penalty": "kl",
+  "lam": 0.95,
+  "learning_rate": 1e-05,
+  "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:e418640b4bc77f01d6e55b2f9b0929625114ca5d4d621637ae8fd966355d2a55
+size 6037

tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "is_local": false,
+  "model_max_length": 1024,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}