Text Generation
Transformers
TensorBoard
Safetensors
mistral
alignment-handbook
trl
orpo
Generated from Trainer
conversational
text-generation-inference
Instructions to use orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch") model = AutoModelForCausalLM.from_pretrained("orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch
- SGLang
How to use orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch with Docker Model Runner:
docker model run hf.co/orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch
Download trainer_state.json from orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch: direct link, hf CLI and curl.
- Browser
- Download file 28.8 kB
-
https://huggingface.co/orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch/resolve/main/trainer_state.json
- Command line
-
hf download hf://orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/orpo-explorers/kaist-mistral-orpo-OHP-15k-Stratified-1-beta-0.2-1epoch-capybara-2epoch/resolve/main/trainer_state.json
28.8 kB
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9954853273137698, | |
| "eval_steps": 500, | |
| "global_step": 442, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 4.9936877763590664e-06, | |
| "log_odds_chosen": 0.2890412509441376, | |
| "log_odds_ratio": -0.6332851648330688, | |
| "logits/chosen": -3.1468911170959473, | |
| "logits/rejected": -3.138887643814087, | |
| "logps/chosen": -0.8719667196273804, | |
| "logps/rejected": -1.0382100343704224, | |
| "loss": 0.6482, | |
| "nll_loss": 0.5332359075546265, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.17439337074756622, | |
| "rewards/margins": 0.033248625695705414, | |
| "rewards/rejected": -0.20764198899269104, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 4.9747829807701e-06, | |
| "log_odds_chosen": 0.3164082169532776, | |
| "log_odds_ratio": -0.6052478551864624, | |
| "logits/chosen": -3.0846054553985596, | |
| "logits/rejected": -3.0846097469329834, | |
| "logps/chosen": -0.8607593774795532, | |
| "logps/rejected": -1.0491193532943726, | |
| "loss": 0.6131, | |
| "nll_loss": 0.4854271411895752, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.17215189337730408, | |
| "rewards/margins": 0.03767203167080879, | |
| "rewards/rejected": -0.20982392132282257, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.943381078271214e-06, | |
| "log_odds_chosen": 0.3783247768878937, | |
| "log_odds_ratio": -0.6129769086837769, | |
| "logits/chosen": -3.084336996078491, | |
| "logits/rejected": -3.0786914825439453, | |
| "logps/chosen": -0.8424398303031921, | |
| "logps/rejected": -1.0533756017684937, | |
| "loss": 0.602, | |
| "nll_loss": 0.4758261740207672, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.16848795115947723, | |
| "rewards/margins": 0.0421871617436409, | |
| "rewards/rejected": -0.21067512035369873, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 3.0, | |
| "learning_rate": 4.89964064152747e-06, | |
| "log_odds_chosen": 0.3220486640930176, | |
| "log_odds_ratio": -0.6127817630767822, | |
| "logits/chosen": -3.073997974395752, | |
| "logits/rejected": -3.069819688796997, | |
| "logps/chosen": -0.8541064262390137, | |
| "logps/rejected": -1.0315220355987549, | |
| "loss": 0.5993, | |
| "nll_loss": 0.47563648223876953, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.17082129418849945, | |
| "rewards/margins": 0.03548307344317436, | |
| "rewards/rejected": -0.2063043862581253, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 4.84378255007397e-06, | |
| "log_odds_chosen": 0.3029581606388092, | |
| "log_odds_ratio": -0.6211820840835571, | |
| "logits/chosen": -3.0848793983459473, | |
| "logits/rejected": -3.0841445922851562, | |
| "logps/chosen": -0.8334375619888306, | |
| "logps/rejected": -1.0042452812194824, | |
| "loss": 0.5913, | |
| "nll_loss": 0.4705049991607666, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1666875034570694, | |
| "rewards/margins": 0.03416154906153679, | |
| "rewards/rejected": -0.20084905624389648, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 4.7760888749230414e-06, | |
| "log_odds_chosen": 0.4831446707248688, | |
| "log_odds_ratio": -0.5493755340576172, | |
| "logits/chosen": -3.095670461654663, | |
| "logits/rejected": -3.0957283973693848, | |
| "logps/chosen": -0.8023236989974976, | |
| "logps/rejected": -1.0722805261611938, | |
| "loss": 0.5964, | |
| "nll_loss": 0.4757974147796631, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.16046474874019623, | |
| "rewards/margins": 0.05399137735366821, | |
| "rewards/rejected": -0.21445612609386444, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 4.696901454167989e-06, | |
| "log_odds_chosen": 0.3349444270133972, | |
| "log_odds_ratio": -0.6006864309310913, | |
| "logits/chosen": -3.0792784690856934, | |
| "logits/rejected": -3.073970079421997, | |
| "logps/chosen": -0.8161932229995728, | |
| "logps/rejected": -0.9950379133224487, | |
| "loss": 0.6022, | |
| "nll_loss": 0.48054131865501404, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.16323864459991455, | |
| "rewards/margins": 0.035768941044807434, | |
| "rewards/rejected": -0.19900760054588318, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 4.6066201667762944e-06, | |
| "log_odds_chosen": 0.46497654914855957, | |
| "log_odds_ratio": -0.5399398803710938, | |
| "logits/chosen": -3.1036829948425293, | |
| "logits/rejected": -3.099396228790283, | |
| "logps/chosen": -0.8102518916130066, | |
| "logps/rejected": -1.0698442459106445, | |
| "loss": 0.5611, | |
| "nll_loss": 0.47170430421829224, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.16205039620399475, | |
| "rewards/margins": 0.05191843584179878, | |
| "rewards/rejected": -0.21396884322166443, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 4.505700913289246e-06, | |
| "log_odds_chosen": 0.385485976934433, | |
| "log_odds_ratio": -0.5975229144096375, | |
| "logits/chosen": -3.096364974975586, | |
| "logits/rejected": -3.085583209991455, | |
| "logps/chosen": -0.8479864001274109, | |
| "logps/rejected": -1.076240062713623, | |
| "loss": 0.6009, | |
| "nll_loss": 0.47817081212997437, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.1695972979068756, | |
| "rewards/margins": 0.04565075412392616, | |
| "rewards/rejected": -0.21524803340435028, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.3946533136249926e-06, | |
| "log_odds_chosen": 0.43778282403945923, | |
| "log_odds_ratio": -0.5804494619369507, | |
| "logits/chosen": -3.0998694896698, | |
| "logits/rejected": -3.0987915992736816, | |
| "logps/chosen": -0.8241409063339233, | |
| "logps/rejected": -1.0574108362197876, | |
| "loss": 0.5988, | |
| "nll_loss": 0.4965757429599762, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.16482816636562347, | |
| "rewards/margins": 0.046653974801301956, | |
| "rewards/rejected": -0.21148213744163513, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 4.274038133610629e-06, | |
| "log_odds_chosen": 0.46187520027160645, | |
| "log_odds_ratio": -0.5494967103004456, | |
| "logits/chosen": -3.1161484718322754, | |
| "logits/rejected": -3.114893913269043, | |
| "logps/chosen": -0.7468121647834778, | |
| "logps/rejected": -0.9928558468818665, | |
| "loss": 0.5811, | |
| "nll_loss": 0.4424869120121002, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.14936241507530212, | |
| "rewards/margins": 0.04920873045921326, | |
| "rewards/rejected": -0.19857117533683777, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.375, | |
| "learning_rate": 4.1444644532387485e-06, | |
| "log_odds_chosen": 0.4036913812160492, | |
| "log_odds_ratio": -0.6148982048034668, | |
| "logits/chosen": -3.0943081378936768, | |
| "logits/rejected": -3.089369297027588, | |
| "logps/chosen": -0.8185051083564758, | |
| "logps/rejected": -1.0283105373382568, | |
| "loss": 0.5743, | |
| "nll_loss": 0.4471573233604431, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.16370102763175964, | |
| "rewards/margins": 0.04196108505129814, | |
| "rewards/rejected": -0.2056621015071869, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 4.006586590948141e-06, | |
| "log_odds_chosen": 0.38943806290626526, | |
| "log_odds_ratio": -0.5949686169624329, | |
| "logits/chosen": -3.105708360671997, | |
| "logits/rejected": -3.1006293296813965, | |
| "logps/chosen": -0.8150092363357544, | |
| "logps/rejected": -1.0240681171417236, | |
| "loss": 0.5787, | |
| "nll_loss": 0.46157145500183105, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.16300183534622192, | |
| "rewards/margins": 0.041811779141426086, | |
| "rewards/rejected": -0.20481359958648682, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 3.861100799460336e-06, | |
| "log_odds_chosen": 0.5897554159164429, | |
| "log_odds_ratio": -0.542380690574646, | |
| "logits/chosen": -3.100862503051758, | |
| "logits/rejected": -3.0946009159088135, | |
| "logps/chosen": -0.7791546583175659, | |
| "logps/rejected": -1.0703848600387573, | |
| "loss": 0.5747, | |
| "nll_loss": 0.5026167631149292, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.15583094954490662, | |
| "rewards/margins": 0.05824603885412216, | |
| "rewards/rejected": -0.214076966047287, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 3.7087417498572946e-06, | |
| "log_odds_chosen": 0.4274469316005707, | |
| "log_odds_ratio": -0.5712376236915588, | |
| "logits/chosen": -3.102602005004883, | |
| "logits/rejected": -3.1062116622924805, | |
| "logps/chosen": -0.8305424451828003, | |
| "logps/rejected": -1.077083945274353, | |
| "loss": 0.5692, | |
| "nll_loss": 0.4742380678653717, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.16610850393772125, | |
| "rewards/margins": 0.04930829256772995, | |
| "rewards/rejected": -0.2154167890548706, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 3.550278821654866e-06, | |
| "log_odds_chosen": 0.34345707297325134, | |
| "log_odds_ratio": -0.6129624247550964, | |
| "logits/chosen": -3.095494508743286, | |
| "logits/rejected": -3.0863471031188965, | |
| "logps/chosen": -0.8034600019454956, | |
| "logps/rejected": -0.9921743273735046, | |
| "loss": 0.587, | |
| "nll_loss": 0.4692765176296234, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.1606920212507248, | |
| "rewards/margins": 0.037742845714092255, | |
| "rewards/rejected": -0.19843485951423645, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3.386512217606339e-06, | |
| "log_odds_chosen": 0.3175373375415802, | |
| "log_odds_ratio": -0.6278958320617676, | |
| "logits/chosen": -3.1244797706604004, | |
| "logits/rejected": -3.1204514503479004, | |
| "logps/chosen": -0.7932506799697876, | |
| "logps/rejected": -0.9527732133865356, | |
| "loss": 0.6102, | |
| "nll_loss": 0.472064733505249, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.15865013003349304, | |
| "rewards/margins": 0.031904518604278564, | |
| "rewards/rejected": -0.1905546486377716, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 3.218268922855452e-06, | |
| "log_odds_chosen": 0.42049679160118103, | |
| "log_odds_ratio": -0.5808514356613159, | |
| "logits/chosen": -3.1028733253479004, | |
| "logits/rejected": -3.0980591773986816, | |
| "logps/chosen": -0.7747179269790649, | |
| "logps/rejected": -0.9938864707946777, | |
| "loss": 0.5825, | |
| "nll_loss": 0.44331270456314087, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.154943585395813, | |
| "rewards/margins": 0.04383372142910957, | |
| "rewards/rejected": -0.19877730309963226, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 3.046398528844248e-06, | |
| "log_odds_chosen": 0.5076274275779724, | |
| "log_odds_ratio": -0.5742210745811462, | |
| "logits/chosen": -3.104665517807007, | |
| "logits/rejected": -3.095426082611084, | |
| "logps/chosen": -0.776441216468811, | |
| "logps/rejected": -1.0134727954864502, | |
| "loss": 0.5712, | |
| "nll_loss": 0.4288255274295807, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.15528824925422668, | |
| "rewards/margins": 0.04740630090236664, | |
| "rewards/rejected": -0.20269456505775452, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 2.871768943064129e-06, | |
| "log_odds_chosen": 0.3015575408935547, | |
| "log_odds_ratio": -0.6296088695526123, | |
| "logits/chosen": -3.0799665451049805, | |
| "logits/rejected": -3.07787823677063, | |
| "logps/chosen": -0.8669988512992859, | |
| "logps/rejected": -1.0326287746429443, | |
| "loss": 0.5744, | |
| "nll_loss": 0.47870683670043945, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.17339977622032166, | |
| "rewards/margins": 0.033125996589660645, | |
| "rewards/rejected": -0.2065257579088211, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 2.695262006314912e-06, | |
| "log_odds_chosen": 0.3043849766254425, | |
| "log_odds_ratio": -0.6296250224113464, | |
| "logits/chosen": -3.0792198181152344, | |
| "logits/rejected": -3.0685715675354004, | |
| "logps/chosen": -0.8420785069465637, | |
| "logps/rejected": -1.0109227895736694, | |
| "loss": 0.5851, | |
| "nll_loss": 0.49167829751968384, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.16841569542884827, | |
| "rewards/margins": 0.03376886993646622, | |
| "rewards/rejected": -0.20218458771705627, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.390625, | |
| "learning_rate": 2.517769039603744e-06, | |
| "log_odds_chosen": 0.47437596321105957, | |
| "log_odds_ratio": -0.5693033933639526, | |
| "logits/chosen": -3.0923125743865967, | |
| "logits/rejected": -3.08634614944458, | |
| "logps/chosen": -0.7851861119270325, | |
| "logps/rejected": -1.0160605907440186, | |
| "loss": 0.6031, | |
| "nll_loss": 0.4828670620918274, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.15703722834587097, | |
| "rewards/margins": 0.04617489501833916, | |
| "rewards/rejected": -0.20321211218833923, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.3401863431710864e-06, | |
| "log_odds_chosen": 0.8126243352890015, | |
| "log_odds_ratio": -0.4562348425388336, | |
| "logits/chosen": -3.0906317234039307, | |
| "logits/rejected": -3.08894681930542, | |
| "logps/chosen": -0.6749258041381836, | |
| "logps/rejected": -1.0676677227020264, | |
| "loss": 0.5063, | |
| "nll_loss": 0.4003027081489563, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.13498517870903015, | |
| "rewards/margins": 0.0785483792424202, | |
| "rewards/rejected": -0.21353355050086975, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 2.163410670372652e-06, | |
| "log_odds_chosen": 0.8564330339431763, | |
| "log_odds_ratio": -0.4268684387207031, | |
| "logits/chosen": -3.0698952674865723, | |
| "logits/rejected": -3.070732831954956, | |
| "logps/chosen": -0.7170256972312927, | |
| "logps/rejected": -1.1589481830596924, | |
| "loss": 0.4975, | |
| "nll_loss": 0.45029813051223755, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.14340513944625854, | |
| "rewards/margins": 0.08838452398777008, | |
| "rewards/rejected": -0.23178966343402863, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 1.9883346992732256e-06, | |
| "log_odds_chosen": 1.0222368240356445, | |
| "log_odds_ratio": -0.36799290776252747, | |
| "logits/chosen": -3.055107593536377, | |
| "logits/rejected": -3.045011281967163, | |
| "logps/chosen": -0.6467097997665405, | |
| "logps/rejected": -1.2033334970474243, | |
| "loss": 0.4936, | |
| "nll_loss": 0.3834058940410614, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.1293419599533081, | |
| "rewards/margins": 0.1113247275352478, | |
| "rewards/rejected": -0.24066667258739471, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 1.8158425248197931e-06, | |
| "log_odds_chosen": 0.9676607251167297, | |
| "log_odds_ratio": -0.3842445909976959, | |
| "logits/chosen": -3.078850269317627, | |
| "logits/rejected": -3.0772275924682617, | |
| "logps/chosen": -0.6771007180213928, | |
| "logps/rejected": -1.177585482597351, | |
| "loss": 0.5028, | |
| "nll_loss": 0.44024211168289185, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.13542014360427856, | |
| "rewards/margins": 0.10009696334600449, | |
| "rewards/rejected": -0.23551711440086365, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 1.6468051943575242e-06, | |
| "log_odds_chosen": 0.9456725120544434, | |
| "log_odds_ratio": -0.39273351430892944, | |
| "logits/chosen": -3.085444450378418, | |
| "logits/rejected": -3.08333420753479, | |
| "logps/chosen": -0.6596750020980835, | |
| "logps/rejected": -1.1495213508605957, | |
| "loss": 0.4758, | |
| "nll_loss": 0.3547888398170471, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.1319349855184555, | |
| "rewards/margins": 0.09796925634145737, | |
| "rewards/rejected": -0.22990426421165466, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 1.482076309033254e-06, | |
| "log_odds_chosen": 0.8887540102005005, | |
| "log_odds_ratio": -0.43248239159584045, | |
| "logits/chosen": -3.0775530338287354, | |
| "logits/rejected": -3.0756869316101074, | |
| "logps/chosen": -0.6567140817642212, | |
| "logps/rejected": -1.1050218343734741, | |
| "loss": 0.4872, | |
| "nll_loss": 0.3975124657154083, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.1313427984714508, | |
| "rewards/margins": 0.08966155350208282, | |
| "rewards/rejected": -0.22100436687469482, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 1.3224877132984131e-06, | |
| "log_odds_chosen": 1.054574728012085, | |
| "log_odds_ratio": -0.3588215708732605, | |
| "logits/chosen": -3.103698253631592, | |
| "logits/rejected": -3.107422351837158, | |
| "logps/chosen": -0.6697841882705688, | |
| "logps/rejected": -1.2304431200027466, | |
| "loss": 0.4864, | |
| "nll_loss": 0.4196816384792328, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -0.13395683467388153, | |
| "rewards/margins": 0.11213177442550659, | |
| "rewards/rejected": -0.24608862400054932, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 1.1688452942784592e-06, | |
| "log_odds_chosen": 1.018751621246338, | |
| "log_odds_ratio": -0.3791121542453766, | |
| "logits/chosen": -3.0898470878601074, | |
| "logits/rejected": -3.0825257301330566, | |
| "logps/chosen": -0.6466778516769409, | |
| "logps/rejected": -1.1818552017211914, | |
| "loss": 0.4914, | |
| "nll_loss": 0.3884134292602539, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.12933556735515594, | |
| "rewards/margins": 0.10703550279140472, | |
| "rewards/rejected": -0.23637107014656067, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 1.021924912221062e-06, | |
| "log_odds_chosen": 0.8282191157341003, | |
| "log_odds_ratio": -0.42912721633911133, | |
| "logits/chosen": -3.1064085960388184, | |
| "logits/rejected": -3.105886697769165, | |
| "logps/chosen": -0.669070839881897, | |
| "logps/rejected": -1.1031153202056885, | |
| "loss": 0.4969, | |
| "nll_loss": 0.4509221017360687, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.13381414115428925, | |
| "rewards/margins": 0.08680891990661621, | |
| "rewards/rejected": -0.22062306106090546, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.824684825733865e-07, | |
| "log_odds_chosen": 1.0580097436904907, | |
| "log_odds_ratio": -0.38433724641799927, | |
| "logits/chosen": -3.081042528152466, | |
| "logits/rejected": -3.075192451477051, | |
| "logps/chosen": -0.6307107210159302, | |
| "logps/rejected": -1.1325790882110596, | |
| "loss": 0.493, | |
| "nll_loss": 0.4108586311340332, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.12614212930202484, | |
| "rewards/margins": 0.10037368535995483, | |
| "rewards/rejected": -0.22651581466197968, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 7.51180229473116e-07, | |
| "log_odds_chosen": 0.9154760241508484, | |
| "log_odds_ratio": -0.415397584438324, | |
| "logits/chosen": -3.0937957763671875, | |
| "logits/rejected": -3.0928263664245605, | |
| "logps/chosen": -0.6775897741317749, | |
| "logps/rejected": -1.1316546201705933, | |
| "loss": 0.5044, | |
| "nll_loss": 0.4353519380092621, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.13551795482635498, | |
| "rewards/margins": 0.09081296622753143, | |
| "rewards/rejected": -0.22633090615272522, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 6.28723129572247e-07, | |
| "log_odds_chosen": 1.139094352722168, | |
| "log_odds_ratio": -0.351654052734375, | |
| "logits/chosen": -3.077972888946533, | |
| "logits/rejected": -3.0774223804473877, | |
| "logps/chosen": -0.6138228178024292, | |
| "logps/rejected": -1.158979058265686, | |
| "loss": 0.4728, | |
| "nll_loss": 0.3857228457927704, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.12276456505060196, | |
| "rewards/margins": 0.10903123766183853, | |
| "rewards/rejected": -0.2317957878112793, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 5.157155641515766e-07, | |
| "log_odds_chosen": 1.0156279802322388, | |
| "log_odds_ratio": -0.3680209219455719, | |
| "logits/chosen": -3.093818426132202, | |
| "logits/rejected": -3.095491409301758, | |
| "logps/chosen": -0.6614006161689758, | |
| "logps/rejected": -1.1964319944381714, | |
| "loss": 0.5024, | |
| "nll_loss": 0.4554772973060608, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.13228009641170502, | |
| "rewards/margins": 0.10700628906488419, | |
| "rewards/rejected": -0.2392864227294922, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 4.127281964319446e-07, | |
| "log_odds_chosen": 1.1072419881820679, | |
| "log_odds_ratio": -0.365578830242157, | |
| "logits/chosen": -3.0632922649383545, | |
| "logits/rejected": -3.053469657897949, | |
| "logps/chosen": -0.6406615376472473, | |
| "logps/rejected": -1.1723295450210571, | |
| "loss": 0.4717, | |
| "nll_loss": 0.39158958196640015, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.12813231348991394, | |
| "rewards/margins": 0.10633359849452972, | |
| "rewards/rejected": -0.23446591198444366, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 3.202810898511424e-07, | |
| "log_odds_chosen": 1.0118036270141602, | |
| "log_odds_ratio": -0.3721558451652527, | |
| "logits/chosen": -3.0760838985443115, | |
| "logits/rejected": -3.0786654949188232, | |
| "logps/chosen": -0.6295793652534485, | |
| "logps/rejected": -1.1554322242736816, | |
| "loss": 0.4954, | |
| "nll_loss": 0.3821503221988678, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.12591585516929626, | |
| "rewards/margins": 0.1051705852150917, | |
| "rewards/rejected": -0.23108646273612976, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 2.388410818585263e-07, | |
| "log_odds_chosen": 1.0309429168701172, | |
| "log_odds_ratio": -0.3757808804512024, | |
| "logits/chosen": -3.0944666862487793, | |
| "logits/rejected": -3.102954149246216, | |
| "logps/chosen": -0.6560705304145813, | |
| "logps/rejected": -1.1918388605117798, | |
| "loss": 0.4848, | |
| "nll_loss": 0.4180088937282562, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.13121411204338074, | |
| "rewards/margins": 0.10715366899967194, | |
| "rewards/rejected": -0.23836776614189148, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 1.6881942648911077e-07, | |
| "log_odds_chosen": 1.045115351676941, | |
| "log_odds_ratio": -0.3773229420185089, | |
| "logits/chosen": -3.099276304244995, | |
| "logits/rejected": -3.0971086025238037, | |
| "logps/chosen": -0.6689329743385315, | |
| "logps/rejected": -1.209533452987671, | |
| "loss": 0.4865, | |
| "nll_loss": 0.4016016125679016, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.13378658890724182, | |
| "rewards/margins": 0.10812009871006012, | |
| "rewards/rejected": -0.24190667271614075, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 1.1056971762161584e-07, | |
| "log_odds_chosen": 0.8860037922859192, | |
| "log_odds_ratio": -0.4059659540653229, | |
| "logits/chosen": -3.0970611572265625, | |
| "logits/rejected": -3.0827975273132324, | |
| "logps/chosen": -0.6797454357147217, | |
| "logps/rejected": -1.1361665725708008, | |
| "loss": 0.4883, | |
| "nll_loss": 0.4087887406349182, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.13594909012317657, | |
| "rewards/margins": 0.09128421545028687, | |
| "rewards/rejected": -0.22723329067230225, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 6.438610340755464e-08, | |
| "log_odds_chosen": 0.9491767883300781, | |
| "log_odds_ratio": -0.3955945372581482, | |
| "logits/chosen": -3.095716953277588, | |
| "logits/rejected": -3.0901029109954834, | |
| "logps/chosen": -0.6622263193130493, | |
| "logps/rejected": -1.1627901792526245, | |
| "loss": 0.4715, | |
| "nll_loss": 0.4081706404685974, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.13244526088237762, | |
| "rewards/margins": 0.10011278092861176, | |
| "rewards/rejected": -0.23255804181098938, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 3.050180088809973e-08, | |
| "log_odds_chosen": 0.89338219165802, | |
| "log_odds_ratio": -0.41199594736099243, | |
| "logits/chosen": -3.078094482421875, | |
| "logits/rejected": -3.074589252471924, | |
| "logps/chosen": -0.6442387700080872, | |
| "logps/rejected": -1.103659987449646, | |
| "loss": 0.4976, | |
| "nll_loss": 0.40062204003334045, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.12884774804115295, | |
| "rewards/margins": 0.09188426285982132, | |
| "rewards/rejected": -0.22073200345039368, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 9.087918299586772e-09, | |
| "log_odds_chosen": 0.9813147783279419, | |
| "log_odds_ratio": -0.4028463363647461, | |
| "logits/chosen": -3.079704523086548, | |
| "logits/rejected": -3.0714404582977295, | |
| "logps/chosen": -0.6575345993041992, | |
| "logps/rejected": -1.1326944828033447, | |
| "loss": 0.4832, | |
| "nll_loss": 0.39662784337997437, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.13150693476200104, | |
| "rewards/margins": 0.0950319841504097, | |
| "rewards/rejected": -0.22653889656066895, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.525910147516131e-10, | |
| "log_odds_chosen": 0.966279149055481, | |
| "log_odds_ratio": -0.4019397795200348, | |
| "logits/chosen": -3.0766844749450684, | |
| "logits/rejected": -3.0788207054138184, | |
| "logps/chosen": -0.6589742302894592, | |
| "logps/rejected": -1.1595604419708252, | |
| "loss": 0.48, | |
| "nll_loss": 0.38374659419059753, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.13179484009742737, | |
| "rewards/margins": 0.10011725127696991, | |
| "rewards/rejected": -0.23191209137439728, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 442, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5399103337283587, | |
| "train_runtime": 8872.0808, | |
| "train_samples_per_second": 3.191, | |
| "train_steps_per_second": 0.05 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 442, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |