Add 11 files

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+training-metrics.png filter=lfs diff=lfs merge=lfs -text
+win-rates.png filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,102 @@
+---
+base_model: meta-llama/Llama-3.2-3B-Instruct
+library_name: peft
+license: other
+tags:
+  - lora
+  - peft
+  - adapter
+  - adaption
+---
+
+# adaption_africa_math_code_qa
+
+## Model Training
+
+A LORA adapter for `meta-llama/Llama-3.2-3B-Instruct`. This model was trained with SFT using [Adaption](https://adaptionlabs.ai)'s AutoScientist on the africa_math_code_qa dataset.
+
+
+![Training metrics](training-metrics.png)
+
+### AutoScientist Config
+
+```json
+{
+  "job_id": "d70be010-a89f-464a-a6c9-928fae8e0d79",
+  "training_experiment_id": "2904a8f8-eaab-4afa-81cb-1a5473972fd3",
+  "original_model_name": "meta-llama/Llama-3.2-3B-Instruct",
+  "trained_model_name": "adaption_africa_math_code_qa",
+  "training_method": "sft",
+  "training_type": "lora",
+  "data_format": "chat",
+  "hyperparams": {
+    "lora": "true",
+    "lora_r": 16,
+    "n_evals": 5,
+    "n_epochs": 1,
+    "batch_size": "max",
+    "lora_alpha": 32,
+    "lora_dropout": 0,
+    "min_lr_ratio": 0.1,
+    "warmup_ratio": 0.1,
+    "weight_decay": 0,
+    "learning_rate": 0.00001,
+    "max_grad_norm": 2,
+    "base_model_size": "3B",
+    "train_on_inputs": "false",
+    "training_method": "sft",
+    "lr_scheduler_type": "cosine",
+    "scheduler_num_cycles": 0.5,
+    "lora_trainable_modules": "all-linear"
+  }
+}
+```
+
+## Training Data
+
+The model was trained on 27,523 rows of adapted data with the following domain distribution: code (31%), math (20%), agriculture (18%), personal-finance (10%), geography (4%), technology (4%), science (4%), governance (3%), corporate-business (2%), how-to (1%), travel (1%), architecture-design (1%), legal (0%), language (0%), education (0%), marketing (0%), data-analysis-visualization (0%).
+
+## Model Evaluation
+
+The model was evaluated on an in-distribution held-out test set as well as a broader domain-specific test set to measure generalization.
+
+
+![Win rates](win-rates.png)
+
+| Domain | Win rate vs. base model |
+| --- | --- |
+| general | 54% |
+
+## How to use
+
+```bash
+pip install torch transformers peft
+```
+
+```python
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from peft import PeftModel
+
+BASE = "meta-llama/Llama-3.2-3B-Instruct"
+ADAPTER = "<this-repo-id>"
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+dtype = torch.float32 if device == "cpu" else torch.bfloat16
+
+base = AutoModelForCausalLM.from_pretrained(BASE, dtype=dtype).to(device)
+model = PeftModel.from_pretrained(base, ADAPTER)
+# Optional: merge the LoRA weights into the base for faster inference
+model = model.merge_and_unload()
+model.eval()
+
+tokenizer = AutoTokenizer.from_pretrained(BASE)
+messages = [{"role": "user", "content": "Hello!"}]
+text = tokenizer.apply_chat_template(
+    messages, tokenize=False, add_generation_prompt=True)
+inputs = tokenizer(text, return_tensors="pt").to(device)
+
+with torch.inference_mode():
+    out = model.generate(**inputs, max_new_tokens=512)
+print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
+```

adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "togethercomputer/Meta-Llama-3.2-3B-Instruct-Reference__TOG__FT",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": [],
+  "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": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "down_proj",
+    "o_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3db784d8e952a07aeb7f57d0d02f8dc5992c6b13fa72d864356b65c36817a52
+size 97307544

chat_template.jinja

@@ -0,0 +1,93 @@
+{{- bos_token }}
+{%- if custom_tools is defined %}
+    {%- set tools = custom_tools %}
+{%- endif %}
+{%- if not tools_in_user_message is defined %}
+    {%- set tools_in_user_message = true %}
+{%- endif %}
+{%- if not date_string is defined %}
+    {%- if strftime_now is defined %}
+        {%- set date_string = strftime_now("%d %b %Y") %}
+    {%- else %}
+        {%- set date_string = "26 Jul 2024" %}
+    {%- endif %}
+{%- endif %}
+{%- if not tools is defined %}
+    {%- set tools = none %}
+{%- endif %}
+
+{#- This block extracts the system message, so we can slot it into the right place. #}
+{%- if messages[0]['role'] == 'system' %}
+    {%- set system_message = messages[0]['content']|trim %}
+    {%- set messages = messages[1:] %}
+{%- else %}
+    {%- set system_message = "" %}
+{%- endif %}
+
+{#- System message #}
+{{- "<|start_header_id|>system<|end_header_id|>\n\n" }}
+{%- if tools is not none %}
+    {{- "Environment: ipython\n" }}
+{%- endif %}
+{{- "Cutting Knowledge Date: December 2023\n" }}
+{{- "Today Date: " + date_string + "\n\n" }}
+{%- if tools is not none and not tools_in_user_message %}
+    {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+{%- endif %}
+{{- system_message }}
+{{- "<|eot_id|>" }}
+
+{#- Custom tools are passed in a user message with some extra guidance #}
+{%- if tools_in_user_message and not tools is none %}
+    {#- Extract the first user message so we can plug it in here #}
+    {%- if messages | length != 0 %}
+        {%- set first_user_message = messages[0]['content']|trim %}
+        {%- set messages = messages[1:] %}
+    {%- else %}
+        {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }}
+{%- endif %}
+    {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}}
+    {{- "Given the following functions, please respond with a JSON for a function call " }}
+    {{- "with its proper arguments that best answers the given prompt.\n\n" }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+    {{- first_user_message + "<|eot_id|>"}}
+{%- endif %}
+
+{%- for message in messages %}
+    {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}
+        {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }}
+    {%- elif 'tool_calls' in message %}
+        {%- if not message.tool_calls|length == 1 %}
+            {{- raise_exception("This model only supports single tool-calls at once!") }}
+        {%- endif %}
+        {%- set tool_call = message.tool_calls[0].function %}
+        {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}}
+        {{- '{"name": "' + tool_call.name + '", ' }}
+        {{- '"parameters": ' }}
+        {{- tool_call.arguments | tojson }}
+        {{- "}" }}
+        {{- "<|eot_id|>" }}
+    {%- elif message.role == "tool" or message.role == "ipython" %}
+        {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }}
+        {%- if message.content is mapping or message.content is iterable %}
+            {{- message.content | tojson }}
+        {%- else %}
+            {{- message.content }}
+        {%- endif %}
+        {{- "<|eot_id|>" }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }}
+{%- endif %}

config.json

@@ -0,0 +1,37 @@
+{
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "dtype": "bfloat16",
+  "eos_token_id": 128009,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 3072,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 131072,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 24,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 128009,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_parameters": {
+    "factor": 32.0,
+    "high_freq_factor": 4.0,
+    "low_freq_factor": 1.0,
+    "original_max_position_embeddings": 8192,
+    "rope_theta": 500000.0,
+    "rope_type": "llama3"
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "5.13.0",
+  "use_cache": false,
+  "vocab_size": 128256,
+  "torch_dtype": "bfloat16"
+}

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "bos_token": "<|begin_of_text|>",
+  "eos_token": "<|eot_id|>",
+  "pad_token": "<|eot_id|>"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
+size 17209920

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "is_local": false,
+  "local_files_only": true,
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 131072,
+  "pad_token": "<|eot_id|>",
+  "padding_side": "right",
+  "tokenizer_class": "TokenizersBackend"
+}

trainer_state.json

@@ -0,0 +1,249 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 5,
+  "global_step": 25,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3852219879627228,
+      "learning_rate": 0.0,
+      "loss": 1.6982421875,
+      "step": 1
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3040158450603485,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.77734375,
+      "step": 2
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2961996793746948,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.7578125,
+      "step": 3
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3623664677143097,
+      "learning_rate": 1e-05,
+      "loss": 1.7568359375,
+      "step": 4
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2844073176383972,
+      "learning_rate": 9.954196488464198e-06,
+      "loss": 1.7509765625,
+      "step": 5
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 1.7109375,
+      "eval_runtime": 2.0186,
+      "eval_samples_per_second": 0.991,
+      "eval_steps_per_second": 0.495,
+      "step": 5
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2740273177623749,
+      "learning_rate": 9.81771838126524e-06,
+      "loss": 1.6904296875,
+      "step": 6
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27806246280670166,
+      "learning_rate": 9.593343979095334e-06,
+      "loss": 1.720703125,
+      "step": 7
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30633893609046936,
+      "learning_rate": 9.285640897740316e-06,
+      "loss": 1.7900390625,
+      "step": 8
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2555227279663086,
+      "learning_rate": 8.900873084594164e-06,
+      "loss": 1.4390869140625,
+      "step": 9
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2538067102432251,
+      "learning_rate": 8.446873302753783e-06,
+      "loss": 1.70703125,
+      "step": 10
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.697265625,
+      "eval_runtime": 2.0251,
+      "eval_samples_per_second": 0.988,
+      "eval_steps_per_second": 0.494,
+      "step": 10
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30936306715011597,
+      "learning_rate": 7.932883678550191e-06,
+      "loss": 1.734375,
+      "step": 11
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2755681872367859,
+      "learning_rate": 7.36936755850849e-06,
+      "loss": 1.693359375,
+      "step": 12
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24498508870601654,
+      "learning_rate": 6.767796505786435e-06,
+      "loss": 1.638671875,
+      "step": 13
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30694907903671265,
+      "learning_rate": 6.140416772229785e-06,
+      "loss": 1.6572265625,
+      "step": 14
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.309129536151886,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.6904296875,
+      "step": 15
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.693359375,
+      "eval_runtime": 2.0251,
+      "eval_samples_per_second": 0.988,
+      "eval_steps_per_second": 0.494,
+      "step": 15
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31138136982917786,
+      "learning_rate": 4.859583227770218e-06,
+      "loss": 1.7568359375,
+      "step": 16
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25136393308639526,
+      "learning_rate": 4.232203494213567e-06,
+      "loss": 1.640625,
+      "step": 17
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30998989939689636,
+      "learning_rate": 3.630632441491512e-06,
+      "loss": 1.7490234375,
+      "step": 18
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22774231433868408,
+      "learning_rate": 3.0671163214498127e-06,
+      "loss": 1.6455078125,
+      "step": 19
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38729652762413025,
+      "learning_rate": 2.5531266972462176e-06,
+      "loss": 1.7412109375,
+      "step": 20
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.68359375,
+      "eval_runtime": 2.0237,
+      "eval_samples_per_second": 0.988,
+      "eval_steps_per_second": 0.494,
+      "step": 20
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3755400478839874,
+      "learning_rate": 2.0991269154058387e-06,
+      "loss": 1.703125,
+      "step": 21
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3168826401233673,
+      "learning_rate": 1.7143591022596846e-06,
+      "loss": 1.7587890625,
+      "step": 22
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31902143359184265,
+      "learning_rate": 1.4066560209046673e-06,
+      "loss": 1.794921875,
+      "step": 23
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24556264281272888,
+      "learning_rate": 1.1822816187347625e-06,
+      "loss": 1.4383544921875,
+      "step": 24
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234353542327881,
+      "learning_rate": 1.0458035115358031e-06,
+      "loss": 1.5703125,
+      "step": 25
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.681640625,
+      "eval_runtime": 2.0196,
+      "eval_samples_per_second": 0.99,
+      "eval_steps_per_second": 0.495,
+      "step": 25
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 25,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.4638008283181875e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}