Upload folder using huggingface_hub

README.md

@@ -1,207 +1,62 @@
 ---
+license: apache-2.0
 base_model: mistralai/Mistral-7B-Instruct-v0.3
 library_name: peft
 pipeline_tag: text-generation
+language:
+- en
 tags:
-- base_model:adapter:mistralai/Mistral-7B-Instruct-v0.3
+- crystallography
+- structural-biology
 - lora
-- transformers
+- peft
 ---
 
-# Model Card for Model ID
+# Mistral 7B — crystallography LoRA adapter
 
-<!-- Provide a quick summary of what the model is/does. -->
+LoRA adapter for `mistralai/Mistral-7B-Instruct-v0.3`, fine-tuned on confirmed
+question-and-answer pairs from the CCP4BB, CCP-EM, phenixbb and COOT
+mailing lists.
 
+## Training
 
-
-## Model Details
-
-### Model Description
-
-<!-- Provide a longer summary of what this model is. -->
-
-
-
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-
-### Model Sources [optional]
-
-<!-- Provide the basic links for the model. -->
-
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-
-## Uses
-
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-
-### Direct Use
-
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-
-[More Information Needed]
-
-### Downstream Use [optional]
-
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-
-[More Information Needed]
-
-### Out-of-Scope Use
-
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-
-[More Information Needed]
-
-## Bias, Risks, and Limitations
-
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-
-[More Information Needed]
-
-### Recommendations
-
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-
-## How to Get Started with the Model
-
-Use the code below to get started with the model.
-
-[More Information Needed]
-
-## Training Details
-
-### Training Data
-
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-
-[More Information Needed]
-
-### Training Procedure
-
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-
-#### Preprocessing [optional]
-
-[More Information Needed]
-
-
-#### Training Hyperparameters
-
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-
-#### Speeds, Sizes, Times [optional]
-
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-
-[More Information Needed]
+| | |
+|---|---|
+| Method | LoRA, r=64, α=128, dropout=0.05 |
+| Learning rate | 2.675e-05 |
+| Max length | 2048 tokens |
+| Epochs | 2, early stopping |
+| Seed | 1 |
 
 ## Evaluation
 
-<!-- This section describes the evaluation protocols and provides the results. -->
-
-### Testing Data, Factors & Metrics
-
-#### Testing Data
-
-<!-- This should link to a Dataset Card if possible. -->
-
-[More Information Needed]
-
-#### Factors
-
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-
-[More Information Needed]
-
-#### Metrics
-
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-
-[More Information Needed]
-
-### Results
-
-[More Information Needed]
-
-#### Summary
-
-
-
-## Model Examination [optional]
-
-<!-- Relevant interpretability work for the model goes here -->
-
-[More Information Needed]
-
-## Environmental Impact
-
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-
-## Technical Specifications [optional]
-
-### Model Architecture and Objective
-
-[More Information Needed]
-
-### Compute Infrastructure
-
-[More Information Needed]
-
-#### Hardware
-
-[More Information Needed]
-
-#### Software
-
-[More Information Needed]
-
-## Citation [optional]
-
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-
-**BibTeX:**
-
-[More Information Needed]
-
-**APA:**
-
-[More Information Needed]
-
-## Glossary [optional]
+BERTScore F1 against held-out confirmed answers (433 questions,
+`roberta-large`, `score(prediction, reference)`):
+**0.8367 → 0.8536**.
 
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+BERTScore measures overlap with the reference wording, not factual
+correctness; the paper reports three LLM judges alongside it.
 
-[More Information Needed]
+## Use
 
-## More Information [optional]
+```python
+from peft import PeftModel
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
-[More Information Needed]
+tok = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
+model = AutoModelForCausalLM.from_pretrained(
+    "mistralai/Mistral-7B-Instruct-v0.3", device_map="auto", dtype="bfloat16")
+model = PeftModel.from_pretrained(model, "ealharbi/ccp4bb-mistral-7b-lora")
+```
 
-## Model Card Authors [optional]
+The system prompt used in training was:
 
-[More Information Needed]
+> You are a helpful crystallography assistant. Be concise and precise.
 
-## Model Card Contact
+The training pairs are single-turn, so ask one self-contained question
+at a time; multi-turn prompts fall outside the fine-tuning format.
 
-[More Information Needed]
-### Framework versions
+## Caveat
 
-- PEFT 0.19.1
+Answers are generated and may be wrong. Verify anything consequential
+against the program documentation and the primary literature.

adapter_config.json

@@ -31,12 +31,12 @@
   "revision": null,
   "target_modules": [
     "down_proj",
-    "k_proj",
     "v_proj",
-    "q_proj",
+    "gate_proj",
     "o_proj",
-    "up_proj",
-    "gate_proj"
+    "q_proj",
+    "k_proj",
+    "up_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2525c9e77856b784ce537ca8cdd1c322de2ebdc5198f25ebd5aae7681692176
+oid sha256:5418540686d67f2d7bdd5925e415817d3b1d868865cc71d28e9b3a11f8624f28
 size 671149168

log_history.json

@@ -2,18 +2,18 @@
   "model": "mistralai/Mistral-7B-Instruct-v0.3",
   "seed": 1,
   "epochs_requested": 2,
-  "epoch_reached": 1.296983758700696,
+  "epoch_reached": 1.2962962962962963,
   "global_step": 140,
   "max_steps": 216,
   "early_stopped": true,
-  "best_eval_loss": 2.030566692352295,
+  "best_eval_loss": 1.654082179069519,
   "best_checkpoint": "finetuned_models/mistralai-Mistral-7B-Instruct-v0.3_seed1/checkpoint-110",
   "final_metrics": {
-    "eval_loss": 2.030566692352295,
-    "eval_runtime": 16.7007,
-    "eval_samples_per_second": 25.747,
-    "eval_steps_per_second": 6.467,
-    "epoch": 1.296983758700696
+    "eval_loss": 1.654082179069519,
+    "eval_runtime": 13.1871,
+    "eval_samples_per_second": 32.759,
+    "eval_steps_per_second": 8.19,
+    "epoch": 1.2962962962962963
   },
   "learning_rate": 2.675263789416278e-05,
   "lora": {
@@ -31,167 +31,167 @@
   "early_stopping_patience": 3,
   "log_history": [
     {
-      "eval_loss": 2.141383647918701,
-      "eval_runtime": 11.0521,
-      "eval_samples_per_second": 38.907,
-      "eval_steps_per_second": 9.772,
-      "epoch": 0.09280742459396751,
+      "eval_loss": 1.8488812446594238,
+      "eval_runtime": 9.8489,
+      "eval_samples_per_second": 43.863,
+      "eval_steps_per_second": 10.966,
+      "epoch": 0.09259259259259259,
       "step": 10
     },
     {
-      "eval_loss": 2.103656053543091,
-      "eval_runtime": 10.789,
-      "eval_samples_per_second": 39.855,
-      "eval_steps_per_second": 10.01,
-      "epoch": 0.18561484918793503,
+      "eval_loss": 1.764305591583252,
+      "eval_runtime": 9.8335,
+      "eval_samples_per_second": 43.931,
+      "eval_steps_per_second": 10.983,
+      "epoch": 0.18518518518518517,
       "step": 20
     },
     {
-      "loss": 2.217321014404297,
-      "grad_norm": 1.7167869806289673,
+      "loss": 2.0483358764648436,
+      "grad_norm": 2.1267218589782715,
       "learning_rate": 2.6318280611836455e-05,
-      "epoch": 0.23201856148491878,
+      "epoch": 0.23148148148148148,
       "step": 25
     },
     {
-      "eval_loss": 2.0854732990264893,
-      "eval_runtime": 10.8578,
-      "eval_samples_per_second": 39.603,
-      "eval_steps_per_second": 9.947,
-      "epoch": 0.27842227378190254,
+      "eval_loss": 1.728833556175232,
+      "eval_runtime": 9.9208,
+      "eval_samples_per_second": 43.545,
+      "eval_steps_per_second": 10.886,
+      "epoch": 0.2777777777777778,
       "step": 30
     },
     {
-      "eval_loss": 2.0740013122558594,
-      "eval_runtime": 10.8703,
-      "eval_samples_per_second": 39.557,
-      "eval_steps_per_second": 9.935,
-      "epoch": 0.37122969837587005,
+      "eval_loss": 1.7081001996994019,
+      "eval_runtime": 9.9092,
+      "eval_samples_per_second": 43.596,
+      "eval_steps_per_second": 10.899,
+      "epoch": 0.37037037037037035,
       "step": 40
     },
     {
-      "loss": 1.9985250854492187,
-      "grad_norm": 2.1705639362335205,
+      "loss": 1.7389724731445313,
+      "grad_norm": 2.0094456672668457,
       "learning_rate": 2.4174302619519622e-05,
-      "epoch": 0.46403712296983757,
+      "epoch": 0.46296296296296297,
       "step": 50
     },
     {
-      "eval_loss": 2.0634052753448486,
-      "eval_runtime": 10.8606,
-      "eval_samples_per_second": 39.593,
-      "eval_steps_per_second": 9.944,
-      "epoch": 0.46403712296983757,
+      "eval_loss": 1.6942864656448364,
+      "eval_runtime": 9.9717,
+      "eval_samples_per_second": 43.323,
+      "eval_steps_per_second": 10.831,
+      "epoch": 0.46296296296296297,
       "step": 50
     },
     {
-      "eval_loss": 2.055530071258545,
-      "eval_runtime": 10.9203,
-      "eval_samples_per_second": 39.376,
-      "eval_steps_per_second": 9.89,
-      "epoch": 0.5568445475638051,
+      "eval_loss": 1.6831718683242798,
+      "eval_runtime": 10.0379,
+      "eval_samples_per_second": 43.037,
+      "eval_steps_per_second": 10.759,
+      "epoch": 0.5555555555555556,
       "step": 60
     },
     {
-      "eval_loss": 2.0481231212615967,
-      "eval_runtime": 10.8423,
-      "eval_samples_per_second": 39.66,
-      "eval_steps_per_second": 9.961,
-      "epoch": 0.6496519721577726,
+      "eval_loss": 1.6765761375427246,
+      "eval_runtime": 10.0079,
+      "eval_samples_per_second": 43.166,
+      "eval_steps_per_second": 10.791,
+      "epoch": 0.6481481481481481,
       "step": 70
     },
     {
-      "loss": 2.014737548828125,
-      "grad_norm": 1.7425487041473389,
+      "loss": 1.7271536254882813,
+      "grad_norm": 1.6946066617965698,
       "learning_rate": 2.0523325458099818e-05,
-      "epoch": 0.6960556844547564,
+      "epoch": 0.6944444444444444,
       "step": 75
     },
     {
-      "eval_loss": 2.0430397987365723,
-      "eval_runtime": 11.0654,
-      "eval_samples_per_second": 38.86,
-      "eval_steps_per_second": 9.76,
-      "epoch": 0.7424593967517401,
+      "eval_loss": 1.6678999662399292,
+      "eval_runtime": 10.8076,
+      "eval_samples_per_second": 39.972,
+      "eval_steps_per_second": 9.993,
+      "epoch": 0.7407407407407407,
       "step": 80
     },
     {
-      "eval_loss": 2.037876605987549,
-      "eval_runtime": 10.7984,
-      "eval_samples_per_second": 39.821,
-      "eval_steps_per_second": 10.001,
-      "epoch": 0.8352668213457076,
+      "eval_loss": 1.6625776290893555,
+      "eval_runtime": 10.0228,
+      "eval_samples_per_second": 43.102,
+      "eval_steps_per_second": 10.775,
+      "epoch": 0.8333333333333334,
       "step": 90
     },
     {
-      "loss": 1.9247970581054688,
-      "grad_norm": 2.195621967315674,
+      "loss": 1.6924752807617187,
+      "grad_norm": 1.8865303993225098,
       "learning_rate": 1.587489051192989e-05,
-      "epoch": 0.9280742459396751,
+      "epoch": 0.9259259259259259,
       "step": 100
     },
     {
-      "eval_loss": 2.0337796211242676,
-      "eval_runtime": 10.9498,
-      "eval_samples_per_second": 39.27,
-      "eval_steps_per_second": 9.863,
-      "epoch": 0.9280742459396751,
+      "eval_loss": 1.6575430631637573,
+      "eval_runtime": 9.9455,
+      "eval_samples_per_second": 43.437,
+      "eval_steps_per_second": 10.859,
+      "epoch": 0.9259259259259259,
       "step": 100
     },
     {
-      "eval_loss": 2.030566692352295,
-      "eval_runtime": 10.9028,
-      "eval_samples_per_second": 39.439,
-      "eval_steps_per_second": 9.906,
-      "epoch": 1.0185614849187936,
+      "eval_loss": 1.654082179069519,
+      "eval_runtime": 9.9907,
+      "eval_samples_per_second": 43.24,
+      "eval_steps_per_second": 10.81,
+      "epoch": 1.0185185185185186,
       "step": 110
     },
     {
-      "eval_loss": 2.0365731716156006,
-      "eval_runtime": 11.0196,
-      "eval_samples_per_second": 39.021,
-      "eval_steps_per_second": 9.801,
-      "epoch": 1.111368909512761,
+      "eval_loss": 1.6632981300354004,
+      "eval_runtime": 9.9243,
+      "eval_samples_per_second": 43.53,
+      "eval_steps_per_second": 10.882,
+      "epoch": 1.1111111111111112,
       "step": 120
     },
     {
-      "loss": 1.867428741455078,
-      "grad_norm": 1.9571967124938965,
+      "loss": 1.5846710205078125,
+      "grad_norm": 1.8140192031860352,
       "learning_rate": 1.0877747382232885e-05,
-      "epoch": 1.1577726218097448,
+      "epoch": 1.1574074074074074,
       "step": 125
     },
     {
-      "eval_loss": 2.041104793548584,
-      "eval_runtime": 10.9251,
-      "eval_samples_per_second": 39.359,
-      "eval_steps_per_second": 9.885,
-      "epoch": 1.2041763341067284,
+      "eval_loss": 1.6580889225006104,
+      "eval_runtime": 9.8547,
+      "eval_samples_per_second": 43.837,
+      "eval_steps_per_second": 10.959,
+      "epoch": 1.2037037037037037,
       "step": 130
     },
     {
-      "eval_loss": 2.038154363632202,
-      "eval_runtime": 11.0005,
-      "eval_samples_per_second": 39.089,
-      "eval_steps_per_second": 9.818,
-      "epoch": 1.296983758700696,
+      "eval_loss": 1.6590867042541504,
+      "eval_runtime": 9.9468,
+      "eval_samples_per_second": 43.431,
+      "eval_steps_per_second": 10.858,
+      "epoch": 1.2962962962962963,
       "step": 140
     },
     {
-      "train_runtime": 488.5924,
-      "train_samples_per_second": 14.106,
-      "train_steps_per_second": 0.442,
-      "total_flos": 7.636717925511987e+16,
-      "train_loss": 1.9583442006792342,
-      "epoch": 1.296983758700696,
+      "train_runtime": 454.6607,
+      "train_samples_per_second": 15.198,
+      "train_steps_per_second": 0.475,
+      "total_flos": 3.648040061632512e+16,
+      "train_loss": 1.7275653975350516,
+      "epoch": 1.2962962962962963,
       "step": 140
     },
     {
-      "eval_loss": 2.030566692352295,
-      "eval_runtime": 16.7007,
-      "eval_samples_per_second": 25.747,
-      "eval_steps_per_second": 6.467,
-      "epoch": 1.296983758700696,
+      "eval_loss": 1.654082179069519,
+      "eval_runtime": 13.1871,
+      "eval_samples_per_second": 32.759,
+      "eval_steps_per_second": 8.19,
+      "epoch": 1.2962962962962963,
       "step": 140
     }
   ]

tokenizer.json

@@ -1,6 +1,11 @@
 {
   "version": "1.0",
-  "truncation": null,
+  "truncation": {
+    "direction": "Right",
+    "max_length": 2048,
+    "strategy": "LongestFirst",
+    "stride": 0
+  },
   "padding": null,
   "added_tokens": [
     {