FurkanNar

GPT-2_Instruct-v0.6

Model & Dataset

Base Model: FurkanNar/GPT-2_Instruct-v0.5
Training Dataset: This version (v0.6) was trained on HuggingFaceH4/ultrachat_200k - A large-scale multi-turn conversation dataset for instruction following
Previous Training: v0.5 was trained on Alpaca, SVAMP, and Dolly-15k datasets

Training Hyperparameters

Table with columns: Parameter, Value
Parameter	Value
Tokenizer	gpt2
Dataset	HuggingFaceH4/ultrachat_200k (10,000 train samples, 1,000 test samples)
Max Sequence Length	512 tokens
Epochs	4
Batch Size	8
Learning Rate	2e-5 (AdamW optimizer)
Max Gradient Norm	1.0 (gradient clipping)
Precision	Mixed Precision (FP16) via torch.cuda.amp autocasting

Training Results

Epoch Summaries

Table with columns: Epoch, Train Loss, Train Perplexity, Val Loss, Val Perplexity
Epoch	Train Loss	Train Perplexity	Val Loss	Val Perplexity
1	2.5944	13.39	2.2924	9.90
2	2.4324	11.39	2.2483	9.47
3	2.3656	10.65	2.2236	9.24
4

Loss Progress

The model shows consistent improvement in both training and validation loss across all epochs. The validation loss decreases from 2.2924 to 2.2063, indicating the model is learning effectively without significant overfitting.

Perplexity Progress

Perplexity follows a similar downward trend, with training perplexity dropping from 13.39 to 10.15 and validation perplexity from 9.90 to 9.08. The narrow gap between training and validation perplexity suggests the model is generalizing well.

Model Artifacts

The trained model is saved locally in the saved_model/ directory:

model.safetensors - Model weights in safetensors format
config.json - Model configuration
generation_config.json - Generation parameters

Training Configuration

The training script includes several optimizations for memory efficiency:

Mixed precision training (FP16) to reduce memory usage
Gradient clipping to prevent exploding gradients
Memory fragmentation reduction via PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True

Inference

The model uses a sophisticated Best-of-N sampling approach with multiple advanced techniques for high-quality generation:

Generation Method

Parallel Batched Generation: Generates N candidates (default N=4) in parallel via GPU batching for efficiency
Length-Normalized Log-Likelihood Scoring: Each candidate is scored using temperature-scaled logits, computing the geometric mean of token probabilities normalized by sequence length to eliminate short-response bias
Calibrated Confidence Selection: Uses a separate calibration temperature (T_calib=0.8) for scoring, independent from generation temperature (T_gen=0.7), allowing separate control over exploration and confidence estimation

Sampling Parameters

Table with columns: Parameter, Value, Description
Parameter	Value	Description
Generation Temperature	0.7	Controls randomness during candidate generation
Calibration Temperature	0.8	Scales logits for confidence scoring
Top-K	40	Limits sampling to top K tokens per position
Top-P (Nucleus)	0.9	Cumulative probability threshold for sampling
Repetition Penalty	1.15	Penalizes repeated tokens to reduce redundancy
Best-of-N

Multi-Turn Conversation

Conversation History: Maintains turn-by-turn history using Dolly-15k/Alpaca format with "Instruction:" and "Response:" tags
Context Management: Automatically truncates earlier conversation turns when context exceeds max_length (512 tokens)
Stop Sequences: Custom stopping criteria prevent over-generation by detecting sequences like "\nInstruction:", "\nResponse:", "\nUser:", etc.

Example Outputs

text

You: How to make a salad --- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.7616 | Geom Mean Prob = 46.7% (171 tokens)   Candidate 2: Log-Likelihood = -0.9515 | Geom Mean Prob = 38.6% (148 tokens)   Candidate 3: Log-Likelihood = -0.9907 | Geom Mean Prob = 37.1% (172 tokens) * Candidate 4: Log-Likelihood = -0.7411 | Geom Mean Prob = 47.7% (172 tokens)-------------------------------------------------------AI: Here's an easy way to make salad:  1. Start with a small bowl of water. 2. Add diced tomatoes and cucumber to it. 3. Mix in a pinch for taste. 4. Add salt or pepper if desired. 5. Pour the dressing over the vegetables before adding them to salads. 6. Serve hot and enjoy! You: How to read a book effectively --- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.9790 | Geom Mean Prob = 37.6% (256 tokens)   Candidate 2: Log-Likelihood = -0.9396 | Geom Mean Prob = 39.1% (256 tokens)   Candidate 3: Log-Likelihood = -1.0409 | Geom Mean Prob = 35.3% (256 tokens) * Candidate 4: Log-Likelihood = -0.8596 | Geom Mean Prob = 42.3% (204 tokens)-------------------------------------------------------AI: Reading books can be a powerful tool for learning about mathematics concepts. Here are some tips on how to read books effectively: - Read a book: You can read books online and look for specific topics that you need help with.- Look up relevant sources: You may find books online that are suitable for reading.

Known Weaknesses

Tight Clusters: The model may generate similar responses across candidates, especially when the training data has limited diversity in certain domains. This can reduce the effectiveness of the Best-of-4 method when candidates are too similar.
GPT-2 Base Knowledge Limitations: Since this model is based on GPT-2 (124M parameters), it has inherent limitations in base knowledge compared to larger models like GPT-3 or GPT-4. The model may:
- Struggle with complex reasoning tasks
- Have limited world knowledge and factual accuracy
- Produce hallucinations or incorrect information
- Have difficulty with specialized domains not well-represented in the training data
Sequence Length Constraint: The model is trained with a max sequence length of 512 tokens, which limits the length and complexity of responses it can generate effectively.
Github Repo
https://github.com/TheOfficialFurkanNar/gpt-2_instruct-v0.6.git

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Container

Run this model inference with full control and performance in your environment.

Learn more

Model Details

Model Provider

FurkanNar

Model Tree

Base

FurkanNar/GPT-2_Instruct-v0.5

Fine-tuned

this model

Input Modalities

Text

Output Modalities

Text

Supported Functionality

Dedicated EndpointsContainer

Explore FriendliAI today

Get started Talk to an engineer

Model & Dataset

Base Model: FurkanNar/GPT-2_Instruct-v0.5
Training Dataset: This version (v0.6) was trained on HuggingFaceH4/ultrachat_200k - A large-scale multi-turn conversation dataset for instruction following
Previous Training: v0.5 was trained on Alpaca, SVAMP, and Dolly-15k datasets

Training Hyperparameters

Table with columns: Parameter, Value
Parameter	Value
Tokenizer	gpt2
Dataset	HuggingFaceH4/ultrachat_200k (10,000 train samples, 1,000 test samples)
Max Sequence Length	512 tokens
Epochs	4
Batch Size	8
Learning Rate	2e-5 (AdamW optimizer)
Max Gradient Norm	1.0 (gradient clipping)
Precision	Mixed Precision (FP16) via torch.cuda.amp autocasting

Training Results

Epoch Summaries

Table with columns: Epoch, Train Loss, Train Perplexity, Val Loss, Val Perplexity
Epoch	Train Loss	Train Perplexity	Val Loss	Val Perplexity
1	2.5944	13.39	2.2924	9.90
2	2.4324	11.39	2.2483	9.47
3	2.3656	10.65	2.2236	9.24
4

Loss Progress

Perplexity Progress

Model Artifacts

The trained model is saved locally in the saved_model/ directory:

model.safetensors - Model weights in safetensors format
config.json - Model configuration
generation_config.json - Generation parameters

Training Configuration

The training script includes several optimizations for memory efficiency:

Mixed precision training (FP16) to reduce memory usage
Gradient clipping to prevent exploding gradients
Memory fragmentation reduction via PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True

Inference

The model uses a sophisticated Best-of-N sampling approach with multiple advanced techniques for high-quality generation:

Generation Method

Parallel Batched Generation: Generates N candidates (default N=4) in parallel via GPU batching for efficiency
Length-Normalized Log-Likelihood Scoring: Each candidate is scored using temperature-scaled logits, computing the geometric mean of token probabilities normalized by sequence length to eliminate short-response bias
Calibrated Confidence Selection: Uses a separate calibration temperature (T_calib=0.8) for scoring, independent from generation temperature (T_gen=0.7), allowing separate control over exploration and confidence estimation

Sampling Parameters

Table with columns: Parameter, Value, Description
Parameter	Value	Description
Generation Temperature	0.7	Controls randomness during candidate generation
Calibration Temperature	0.8	Scales logits for confidence scoring
Top-K	40	Limits sampling to top K tokens per position
Top-P (Nucleus)	0.9	Cumulative probability threshold for sampling
Repetition Penalty	1.15	Penalizes repeated tokens to reduce redundancy
Best-of-N

Multi-Turn Conversation

Conversation History: Maintains turn-by-turn history using Dolly-15k/Alpaca format with "Instruction:" and "Response:" tags
Context Management: Automatically truncates earlier conversation turns when context exceeds max_length (512 tokens)
Stop Sequences: Custom stopping criteria prevent over-generation by detecting sequences like "\nInstruction:", "\nResponse:", "\nUser:", etc.

Example Outputs

text

You: How to make a salad--- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.7616 | Geom Mean Prob = 46.7% (171 tokens)   Candidate 2: Log-Likelihood = -0.9515 | Geom Mean Prob = 38.6% (148 tokens)   Candidate 3: Log-Likelihood = -0.9907 | Geom Mean Prob = 37.1% (172 tokens) * Candidate 4: Log-Likelihood = -0.7411 | Geom Mean Prob = 47.7% (172 tokens)-------------------------------------------------------AI: Here's an easy way to make salad: 1. Start with a small bowl of water. 2. Add diced tomatoes and cucumber to it. 3. Mix in a pinch for taste. 4. Add salt or pepper if desired. 5. Pour the dressing over the vegetables before adding them to salads. 6. Serve hot and enjoy!You: How to read a book effectively--- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.9790 | Geom Mean Prob = 37.6% (256 tokens)   Candidate 2: Log-Likelihood = -0.9396 | Geom Mean Prob = 39.1% (256 tokens)   Candidate 3: Log-Likelihood = -1.0409 | Geom Mean Prob = 35.3% (256 tokens) * Candidate 4: Log-Likelihood = -0.8596 | Geom Mean Prob = 42.3% (204 tokens)-------------------------------------------------------AI: Reading books can be a powerful tool for learning about mathematics concepts. Here are some tips on how to read books effectively:- Read a book: You can read books online and look for specific topics that you need help with.- Look up relevant sources: You may find books online that are suitable for reading.

You: How to make a salad --- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.7616 | Geom Mean Prob = 46.7% (171 tokens)   Candidate 2: Log-Likelihood = -0.9515 | Geom Mean Prob = 38.6% (148 tokens)   Candidate 3: Log-Likelihood = -0.9907 | Geom Mean Prob = 37.1% (172 tokens) * Candidate 4: Log-Likelihood = -0.7411 | Geom Mean Prob = 47.7% (172 tokens)-------------------------------------------------------AI: Here's an easy way to make salad:  1. Start with a small bowl of water. 2. Add diced tomatoes and cucumber to it. 3. Mix in a pinch for taste. 4. Add salt or pepper if desired. 5. Pour the dressing over the vegetables before adding them to salads. 6. Serve hot and enjoy! You: How to read a book effectively --- Best-of-4 Candidate Scores ---   Candidate 1: Log-Likelihood = -0.9790 | Geom Mean Prob = 37.6% (256 tokens)   Candidate 2: Log-Likelihood = -0.9396 | Geom Mean Prob = 39.1% (256 tokens)   Candidate 3: Log-Likelihood = -1.0409 | Geom Mean Prob = 35.3% (256 tokens) * Candidate 4: Log-Likelihood = -0.8596 | Geom Mean Prob = 42.3% (204 tokens)-------------------------------------------------------AI: Reading books can be a powerful tool for learning about mathematics concepts. Here are some tips on how to read books effectively: - Read a book: You can read books online and look for specific topics that you need help with.- Look up relevant sources: You may find books online that are suitable for reading.

Known Weaknesses

Tight Clusters: The model may generate similar responses across candidates, especially when the training data has limited diversity in certain domains. This can reduce the effectiveness of the Best-of-4 method when candidates are too similar.
GPT-2 Base Knowledge Limitations: Since this model is based on GPT-2 (124M parameters), it has inherent limitations in base knowledge compared to larger models like GPT-3 or GPT-4. The model may:
- Struggle with complex reasoning tasks
- Have limited world knowledge and factual accuracy
- Produce hallucinations or incorrect information
- Have difficulty with specialized domains not well-represented in the training data
Sequence Length Constraint: The model is trained with a max sequence length of 512 tokens, which limits the length and complexity of responses it can generate effectively.
Github Repo
https://github.com/TheOfficialFurkanNar/gpt-2_instruct-v0.6.git

GPT-2_Instruct-v0.6

README

Model & Dataset

Training Hyperparameters

Training Results

Epoch Summaries

Loss Progress

Perplexity Progress

Model Artifacts

Training Configuration

Inference

Generation Method

Sampling Parameters

Multi-Turn Conversation

Example Outputs

Known Weaknesses

Github Repo

Explore FriendliAI today

README

Model & Dataset

Training Hyperparameters

Training Results

Epoch Summaries

Loss Progress

Perplexity Progress

Model Artifacts

Training Configuration

Inference

Generation Method

Sampling Parameters

Multi-Turn Conversation

Example Outputs

Known Weaknesses

Github Repo