LLM Name | Goliath LongLORA 120B Rope8 32K Fp16 |
Repository ๐ค | https://huggingface.co/grimulkan/Goliath-longLORA-120b-rope8-32k-fp16 |
Model Size | 120b |
Required VRAM | 235.4 GB |
Updated | 2024-09-07 |
Maintainer | grimulkan |
Model Type | llama |
Model Files | |
Quantization Type | fp16 |
Model Architecture | LlamaForCausalLM |
License | llama2 |
Context Length | 4096 |
Model Max Length | 4096 |
Transformers Version | 4.34.1 |
Tokenizer Class | LlamaTokenizer |
Padding Token | <unk> |
Vocabulary Size | 32000 |
Torch Data Type | float16 |
Model |
Likes |
Downloads |
VRAM |
---|---|---|---|
...gLORA 120B Rope8 32K Fp16 GGUF | 9 | 29 | 43 GB |
...ngLORA 120B Rope8 32K Fp16 AWQ | 4 | 27 | 61 GB |
...gLORA 120B Rope8 32K Fp16 GPTQ | 4 | 25 | 59 GB |
Best Alternatives |
Context / RAM |
Downloads |
Likes |
---|---|---|---|
Miquella 120B 3.0bpw H6 EXL2 | 31K / 44.8 GB | 4 | 10 |
Miquella 120B 8.0bpw H8 EXL2 | 31K / 118.1 GB | 7 | 3 |
Miquella 120B 4.0bpw H6 EXL2 | 31K / 59.4 GB | 4 | 2 |
...t 120B Cat A Llama EXL2 5.5bpw | 8K / 85.3 GB | 339 | 0 |
...t 120B Cat A Llama EXL2 4.5bpw | 8K / 70.3 GB | 4 | 1 |
...RA 120B Rope8 32K 6bpw H8 EXL2 | 4K / 88.7 GB | 4 | 1 |
...egaDolphin 120B 2.9bpw H6 EXL2 | 4K / 44.3 GB | 4 | 3 |
...gaDolphin 120B 2.65bpw H6 EXL2 | 4K / 40.5 GB | 4 | 2 |
...egaDolphin 120B 4.0bpw H6 EXL2 | 4K / 60.8 GB | 4 | 1 |
Goliath 120B 6.0bpw H6 EXL2 | 4K / 88.9 GB | 3 | 1 |
๐ Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! ๐