Phind CodeLlama Models by Phind
About
Code Llama, a family of large language models developed by Meta AI, specializes in code generation and various coding-related tasks, building upon the Llama 2 architecture. Available in sizes ranging from 7B to 70B parameters, these models are designed to accommodate different computational needs. Code Llama is trained on an extensive dataset of code and related texts, enabling it to proficiently generate code in multiple programming languages, explain code snippets, assist with code completion and debugging, and even translate between languages. Meta AI also provides specialized models fine-tuned for tasks like Python programming and instruction following. Released under a permissive license, Code Llama supports both research and commercial applications. Benchmark tests have shown that its performance can match or even surpass other publicly available LLMs.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 8k context, 34B parameters, and structured outputs.
Use when the workload needs 8k context and 34B parameters.
Use when the workload needs 8k context and 34B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Phind CodeLlama 34B V2 | Use when the workload needs 8k context, 34B parameters, and structured outputs. | 2023-08 | 8k context34B parametersstructured outputs | Current |
| Phind CodeLlama 34B V1 | Use when the workload needs 8k context and 34B parameters. | 2023-08 | 8k context34B parameters | Current |
| Phind CodeLlama 34B Python V1 | Use when the workload needs 8k context and 34B parameters. | 2023-08 | 8k context34B parameters | Current |
Release Timeline
1 release groupSpecifications(3 models)
| Model | Released | Context | Parameters | Structured Outputs |
|---|---|---|---|---|
| Phind CodeLlama 34B V2 | 2023-08 | 8k | 34B | Yes |
| Phind CodeLlama 34B V1 | 2023-08 | 8k | 34B | No |
| Phind CodeLlama 34B Python V1 | 2023-08 | 8k | 34B | No |
Available From(3 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Phind CodeLlama 34B V2 | DeepInfra | $0.2 | $0.45 | Serverless |
| Phind CodeLlama 34B V2 | Together AI | $0.8 | $0.8 | Serverless |
| Phind CodeLlama 34B V2 | Fireworks AI | $0.9 | $0.9 | Provisioned |
| Phind CodeLlama 34B V1 | Fireworks AI | $0.9 | $0.9 | Provisioned |
| Phind CodeLlama 34B Python V1 | Fireworks AI | $0.9 | $0.9 | Provisioned |
Frequently Asked Questions
- What is Phind CodeLlama used for?
- Phind CodeLlama is used for coding and structured outputs. The family description and listed model capabilities point to those workloads as the best fit.
- How does Phind CodeLlama compare to StarCoder?
- Phind CodeLlama by Phind is strongest where you need coding, while StarCoder by ServiceNow Research is the closest related family to check for coding. Phind CodeLlama has 3 listed variants and reaches up to 8k context, while StarCoder reaches up to 8k context, so compare the specs and pricing tables before choosing a production model.
- Which Phind CodeLlama model should I use?
- For the lowest listed input price, start with Phind CodeLlama 34B V2 through DeepInfra at $0.2/1M input tokens. For the most capable/latest local choice, evaluate Phind CodeLlama 34B V2 with 8k context and structured outputs.
