End of training

README.md

@@ -1,8 +1,10 @@
 ---
 base_model: BramVanroy/GEITje-7B-ultra
+datasets:
+- BramVanroy/ultra_feedback_dutch
 library_name: peft
-license: cc-by-nc-4.0
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
@@ -17,17 +19,17 @@ should probably proofread and complete it, then remove this comment. -->
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/snoels/huggingface/runs/yng7mdb0)
 # FinGEITje-7B-dpo
 
-This model is a fine-tuned version of [BramVanroy/GEITje-7B-ultra](https://huggingface.co/BramVanroy/GEITje-7B-ultra) on an unknown dataset.
+This model is a fine-tuned version of [/mnt/trained_models/fingeitje](https://huggingface.co//mnt/trained_models/fingeitje) on the BramVanroy/ultra_feedback_dutch dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0279
-- Rewards/chosen: -3.8974
-- Rewards/rejected: -15.9642
-- Rewards/accuracies: 0.9828
-- Rewards/margins: 12.0668
-- Logps/rejected: -1951.9310
-- Logps/chosen: -788.9780
-- Logits/rejected: -1.7371
-- Logits/chosen: -1.8937
+- Rewards/chosen: -3.8986
+- Rewards/rejected: -15.9713
+- Rewards/accuracies: 0.9836
+- Rewards/margins: 12.0727
+- Logps/rejected: -1952.6360
+- Logps/chosen: -789.0983
+- Logits/rejected: -1.7369
+- Logits/chosen: -1.8936
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.999253545658124,
+    "eval_logits/chosen": -1.8935905694961548,
+    "eval_logits/rejected": -1.7368791103363037,
+    "eval_logps/chosen": -789.0983276367188,
+    "eval_logps/rejected": -1952.635986328125,
+    "eval_loss": 0.027867762371897697,
+    "eval_rewards/accuracies": 0.983582079410553,
+    "eval_rewards/chosen": -3.89859676361084,
+    "eval_rewards/margins": 12.0726957321167,
+    "eval_rewards/rejected": -15.971291542053223,
+    "eval_runtime": 830.1728,
+    "eval_samples": 5359,
+    "eval_samples_per_second": 6.455,
+    "eval_steps_per_second": 1.614,
     "total_flos": 0.0,
     "train_loss": 0.0882907345950366,
     "train_runtime": 20220.3954,

config.json

@@ -0,0 +1,42 @@
+{
+  "_name_or_path": "BramVanroy/GEITje-7B-ultra",
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 2,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.42.4",
+  "use_cache": true,
+  "vocab_size": 32000
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.999253545658124,
+    "eval_logits/chosen": -1.8935905694961548,
+    "eval_logits/rejected": -1.7368791103363037,
+    "eval_logps/chosen": -789.0983276367188,
+    "eval_logps/rejected": -1952.635986328125,
+    "eval_loss": 0.027867762371897697,
+    "eval_rewards/accuracies": 0.983582079410553,
+    "eval_rewards/chosen": -3.89859676361084,
+    "eval_rewards/margins": 12.0726957321167,
+    "eval_rewards/rejected": -15.971291542053223,
+    "eval_runtime": 830.1728,
+    "eval_samples": 5359,
+    "eval_samples_per_second": 6.455,
+    "eval_steps_per_second": 1.614
+}

runs/Jul22_11-44-06_supergeldgeit/events.out.tfevents.1721670148.supergeldgeit.16075.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b52be63c11e38470f87e39261c4584e96ced23ba864ab7632a1c061cf603f1ef
+size 828