cput ok for compare

.gitignore

@@ -16,3 +16,6 @@ wandb/
 
 # Training outputs and adapters
 outputs/
+
+# archives and logs
+archives/

README.md

@@ -98,9 +98,15 @@ Key flags:
 - `outputs/` is tracked via Git LFS (`.gitattributes`), so weights can be committed and pushed to the Hub. Run `git lfs install` once, then `git add outputs/...` before committing.
 
 ## Evaluation (inference/compare)
-- Quick smoke test with the saved adapter (edit `lora_dir` inside if you used a different path):
+- Quick smoke test with the saved adapter (edit `lora_dir` or pass flags):
 ```bash
-python evaluation/simple_inference.py
+python evaluation/simple_inference.py \
+  --lora_dir outputs/tinyllama-lora \
+  --device auto \
+  --torch_dtype auto \
+  --max_new_tokens 128 \
+  --temperature 0.7 \
+  --top_p 0.9
 ```
 - Compare base vs. LoRA outputs side-by-side:
 ```bash
@@ -109,6 +115,7 @@ python evaluation/compare_lora.py \
   --lora_dir outputs/tinyllama-lora \
   --prompt "Explain LoRA in one sentence."
 ```
+For CPU or constrained machines, force CPU + fp32 (and add `--offload_dir offload` if using `device_map=auto`):
 ```bash
 python evaluation/compare_lora.py \
   --base_model TinyLlama/TinyLlama-1.1B-Chat-v1.0 \
@@ -116,10 +123,8 @@ python evaluation/compare_lora.py \
   --prompt "Explain LoRA in one sentence." \
   --device cpu \
   --torch_dtype float32
-
-  ```
-
-Optional flags: `--max_new_tokens`, `--temperature`, `--top_p`, `--torch_dtype`.
+```
+Optional flags: `--max_new_tokens`, `--temperature`, `--top_p`, `--torch_dtype`, `--device`, `--offload_dir`.
 
 ## Troubleshooting
 - OOM? Reduce `max_seq_length`, increase `gradient_accumulation_steps`, or switch to a smaller dataset (e.g., use a tiny instruction set like `mlabonne/guanaco-llama2-1k`, or subset your dataset with `--dataset_name your/dataset --max_train_samples 500` in code/script).
@@ -179,4 +184,3 @@ python train_tulu.py \
   --input_field input \
   --output_field output
 
-

archive/outputs-sandy-glade-39/tinyllama-lora/README.md

@@ -0,0 +1,207 @@
+---
+base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:TinyLlama/TinyLlama-1.1B-Chat-v1.0
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## 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]
+
+## 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]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0

archive/outputs-sandy-glade-39/tinyllama-lora/adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

archive/outputs-sandy-glade-39/tinyllama-lora/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31da89b90c5bf37df558866b948363e36921755fd0f57445ddad686da61b518c
+size 72113224

archive/outputs-sandy-glade-39/tinyllama-lora/chat_template.jinja

@@ -0,0 +1,15 @@
+{% for message in messages %}
+{% if message['role'] == 'user' %}
+{{ '<|user|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'system' %}
+{{ '<|system|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'assistant' %}
+{{ '<|assistant|>
+'  + message['content'] + eos_token }}
+{% endif %}
+{% if loop.last and add_generation_prompt %}
+{{ '<|assistant|>' }}
+{% endif %}
+{% endfor %}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/README.md

@@ -0,0 +1,207 @@
+---
+base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:TinyLlama/TinyLlama-1.1B-Chat-v1.0
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## 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]
+
+## 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]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31da89b90c5bf37df558866b948363e36921755fd0f57445ddad686da61b518c
+size 72113224

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/chat_template.jinja

@@ -0,0 +1,15 @@
+{% for message in messages %}
+{% if message['role'] == 'user' %}
+{{ '<|user|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'system' %}
+{{ '<|system|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'assistant' %}
+{{ '<|assistant|>
+'  + message['content'] + eos_token }}
+{% endif %}
+{% if loop.last and add_generation_prompt %}
+{{ '<|assistant|>' }}
+{% endif %}
+{% endfor %}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50f6b39102a3af920ba02801a1c407ef580f9cffd06024522c69a4feca92ff0d
+size 144322618

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5493dbf5bba5741d2b25397f552fc797bb1a5ef8ab1d9d23986917e7fa66c606
+size 13990

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f4d113c6e95c4022178c76328771e94349608fc9f80f5eeae8b4b66f207294b
+size 1064

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/trainer_state.json

@@ -0,0 +1,76 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 63,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15055584907531738,
+      "learning_rate": 0.00017704918032786885,
+      "loss": 1.4055,
+      "step": 10
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11845632642507553,
+      "learning_rate": 0.00014426229508196722,
+      "loss": 1.3037,
+      "step": 20
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14062458276748657,
+      "learning_rate": 0.00011147540983606557,
+      "loss": 1.226,
+      "step": 30
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09463928639888763,
+      "learning_rate": 7.868852459016394e-05,
+      "loss": 1.18,
+      "step": 40
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.10201391577720642,
+      "learning_rate": 4.59016393442623e-05,
+      "loss": 1.1939,
+      "step": 50
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09585852175951004,
+      "learning_rate": 1.3114754098360657e-05,
+      "loss": 1.1497,
+      "step": 60
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 63,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3233386856448000.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

archive/outputs-sandy-glade-39/tinyllama-lora/checkpoint-63/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9eb07f36f9ad0505c1bbe233de6b5d438bfd4e41bafbb6cf5bd00c368046727
+size 5368

archive/outputs-sandy-glade-39/tinyllama-lora/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

archive/outputs-sandy-glade-39/tinyllama-lora/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

archive/outputs-sandy-glade-39/tinyllama-lora/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

archive/outputs-sandy-glade-39/tinyllama-lora/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9eb07f36f9ad0505c1bbe233de6b5d438bfd4e41bafbb6cf5bd00c368046727
+size 5368

evaluation/compare_lora.py

@@ -26,6 +26,11 @@ def parse_args():
         choices=["auto", "cpu", "cuda", "mps"],
         help="Force device map; on CPU use this to keep everything on host.",
     )
+    p.add_argument(
+        "--offload_dir",
+        default=None,
+        help="Optional offload directory when using device_map='auto' on constrained devices.",
+    )
     return p.parse_args()
 
 
@@ -36,7 +41,7 @@ def resolve_dtype(name: str) -> Optional[torch.dtype]:
 
 
 def resolve_device_map(device: str):
-    return {"": "cpu"} if device == "cpu" else "auto"
+    return None if device == "cpu" else "auto"
 
 
 def generate(model, tokenizer, prompt: str, max_new_tokens: int, temperature: float, top_p: float) -> str:
@@ -54,7 +59,8 @@ def generate(model, tokenizer, prompt: str, max_new_tokens: int, temperature: fl
 
 def main():
     args = parse_args()
-    torch_dtype = resolve_dtype(args.torch_dtype) or (torch.float32 if args.device == "cpu" else None)
+    force_cpu = args.device == "cpu"
+    torch_dtype = torch.float32 if force_cpu else resolve_dtype(args.torch_dtype)
     device_map = resolve_device_map(args.device) if args.device != "auto" else "auto"
 
     tokenizer = AutoTokenizer.from_pretrained(args.lora_dir, use_fast=False)
@@ -63,13 +69,19 @@ def main():
         args.base_model,
         device_map=device_map,
         torch_dtype=torch_dtype,
+        offload_folder=args.offload_dir,
     )
     lora_wrapped = AutoModelForCausalLM.from_pretrained(
         args.base_model,
         device_map=device_map,
         torch_dtype=torch_dtype,
+        offload_folder=args.offload_dir,
     )
     lora_wrapped = PeftModel.from_pretrained(lora_wrapped, args.lora_dir)
+    if force_cpu:
+        # Avoid Accelerate dispatch/offload; keep everything on CPU.
+        base_model.to("cpu")
+        lora_wrapped.to("cpu")
 
     base_out = generate(
         base_model,

evaluation/simple_inference.py

@@ -14,7 +14,7 @@ def resolve_dtype(name: str, device: str) -> Optional[torch.dtype]:
 
 
 def resolve_device_map(device: str):
-    return {"": "cpu"} if device == "cpu" else "auto"
+    return None if device == "cpu" else "auto"
 
 
 def parse_args():
@@ -26,6 +26,7 @@ def parse_args():
     p.add_argument("--top_p", type=float, default=0.9)
     p.add_argument("--device", default="auto", choices=["auto", "cpu", "cuda", "mps"])
     p.add_argument("--torch_dtype", default="auto", choices=["auto", "float16", "bfloat16", "float32"])
+    p.add_argument("--offload_dir", default=None, help="Optional offload directory when using device_map='auto'.")
     return p.parse_args()
 
 
@@ -41,6 +42,7 @@ def main():
         base_model,
         device_map=device_map,
         torch_dtype=torch_dtype,
+        offload_folder=args.offload_dir,
     )
     model = PeftModel.from_pretrained(model, args.lora_dir)