Text Generation
Transformers
Safetensors
OpenPeerLLM
English
Arabic
riemann_openpeer
amd
ntk
code
fsdp
rocm
coding
mi325x
decoder
code gen
directml
openpeer
ryzen-ai
reasoning
rewriting
multimodal
paraphrase
prerelease
ntk-trainer
growth-ladder
style-transfer
text-rewriting
riemann-computing
progressive-growth
custom_code
Eval Results (legacy)
Instructions to use OpenPeerAI/My-Name-is-Riemann-Prerelease with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use OpenPeerAI/My-Name-is-Riemann-Prerelease with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="OpenPeerAI/My-Name-is-Riemann-Prerelease", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("OpenPeerAI/My-Name-is-Riemann-Prerelease", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use OpenPeerAI/My-Name-is-Riemann-Prerelease with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "OpenPeerAI/My-Name-is-Riemann-Prerelease" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "OpenPeerAI/My-Name-is-Riemann-Prerelease", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/OpenPeerAI/My-Name-is-Riemann-Prerelease
- SGLang
How to use OpenPeerAI/My-Name-is-Riemann-Prerelease with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "OpenPeerAI/My-Name-is-Riemann-Prerelease" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "OpenPeerAI/My-Name-is-Riemann-Prerelease", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "OpenPeerAI/My-Name-is-Riemann-Prerelease" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "OpenPeerAI/My-Name-is-Riemann-Prerelease", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use OpenPeerAI/My-Name-is-Riemann-Prerelease with Docker Model Runner:
docker model run hf.co/OpenPeerAI/My-Name-is-Riemann-Prerelease
Download training/cloud/verify_before_destroy.ps1 from OpenPeerAI/My-Name-is-Riemann-Prerelease: direct link, hf CLI and curl.
- Browser
- Download file 2.3 kB
-
https://huggingface.co/OpenPeerAI/My-Name-is-Riemann-Prerelease/resolve/main/training/cloud/verify_before_destroy.ps1
- Command line
-
hf download hf://OpenPeerAI/My-Name-is-Riemann-Prerelease/training/cloud/verify_before_destroy.ps1
-
curl -L -o verify_before_destroy.ps1 https://huggingface.co/OpenPeerAI/My-Name-is-Riemann-Prerelease/resolve/main/training/cloud/verify_before_destroy.ps1
2.3 kB
| param( | |
| [string]$NodeIp = (Get-Content "$env:RIEMANN_ROOT\cloud_state\node_ip.txt").Trim(), | |
| [string]$LocalRoot = $env:RIEMANN_ROOT, | |
| [string]$ModelRepo = "" | |
| ) | |
| $ErrorActionPreference = "Continue" | |
| $ok = $true | |
| function Check($label, $cond) { $script:ok = $script:ok -and $cond; "{0} {1}" -f ($(if ($cond) { "PASS" } else { "FAIL" }), $label) } | |
| Check "bootstrap reports ALL DONE on node" ((ssh "root@$NodeIp" "grep -c 'GATE: ALL DONE' /workspace/RiemannModel/logs/bootstrap.log") -match '^[1-9]') | |
| foreach ($s in "stage2_538m", "stage3_1.19b", "stage4_2.65b", "stage5_6.66b", "stage6_12.85b") { | |
| Check "local checkpoint mirrored and verified: $s" (Test-Path "$LocalRoot\checkpoints\$s\.mirrored") | |
| Check "local controller present: $s" ((Test-Path "$LocalRoot\ntk_controllers\$s.pt") -and (Test-Path "$LocalRoot\ntk_controllers\$s.json")) | |
| } | |
| Check "local training logs present" ((Get-ChildItem "$LocalRoot\cloud_logs" -Filter "pretrain_stage6*.log" -ErrorAction SilentlyContinue).Count -ge 1) | |
| if ($ModelRepo) { | |
| $remoteList = ssh "root@$NodeIp" "cd /workspace/RiemannModel/checkpoints/stage6_12.85b && ls -l model-*.safetensors | awk '{print `$5, `$9}'" | |
| $hub = python -c " | |
| from huggingface_hub import HfApi | |
| for f in HfApi().list_repo_tree('$ModelRepo', recursive=True): | |
| s=getattr(f,'size',None) | |
| if s and f.path.endswith('.safetensors'): print(s, f.path.split('/')[-1]) | |
| " | |
| $remoteMap = @{}; foreach ($l in $remoteList) { if ($l -match '^(\d+)\s+(\S+)$') { $remoteMap[$matches[2]] = [long]$matches[1] } } | |
| $hubMap = @{}; foreach ($l in $hub) { if ($l -match '^(\d+)\s+(\S+)$') { $hubMap[$matches[2]] = [long]$matches[1] } } | |
| $shardsMatch = ($remoteMap.Count -gt 0) -and ($remoteMap.Count -eq $hubMap.Count) | |
| foreach ($k in $remoteMap.Keys) { if ($hubMap[$k] -ne $remoteMap[$k]) { $shardsMatch = $false } } | |
| Check "Hub has all $($remoteMap.Count) stage6 shards with matching sizes" $shardsMatch | |
| Check "Hub has stage6 controller" ((python -c "from huggingface_hub import HfApi; print(any(f.path=='ntk_controller/stage6_12.85b.pt' for f in HfApi().list_repo_tree('$ModelRepo', recursive=True)))") -eq "True") | |
| } | |
| "" | |
| if ($ok) { "SAFE TO DESTROY: all archive checks passed." } else { "DO NOT DESTROY: one or more checks failed." } | |