Shanghai AI Lab/Dense

InternLM2 5B

chat
4.5B
Parameters
32K
Context length
1
Benchmarks
4
Quantizations
40K
HF downloads
Architecture
Dense
Released
2024-01-17
Layers
32
KV Heads
8
Head Dim
128
Family
internlm

Quantizations & VRAM

Q4_K_M4.5 bpw
3.0 GB
VRAM required
94%
Quality
Q6_K6.5 bpw
4.1 GB
VRAM required
97%
Quality
Q8_08 bpw
5.0 GB
VRAM required
100%
Quality
FP1616 bpw
9.5 GB
VRAM required
100%
Quality

Benchmarks (1)

HumanEval47.6

GPUs that can run this model

At Q4_K_M quantization. Sorted by minimum VRAM.

Find the best GPU for InternLM2 5B

Build Hardware for InternLM2 5B