Update README.md

README.md

@@ -3,7 +3,11 @@ base_model:
 - CausalLM/35b-beta-long
 ---
 # GGUF quants of CausalLM/35b-beta-long, here I have:
-* IQ4_XS (fits into 24GiB VRAM + 8192 context with q4_1 KV cache, also room for 2048 ubatch)
-* IQ4_NL (fits into 24GiB VRAM + 8192 context with q4_1 KV cache)
-* Q4_K_M (fits into 24GiB VRAM + 6144 context with q4_1 KV cache, also good for CPU inference on E5-26xx v3/v4)
-* Q8_0 (probably isn't practical for anything unless you have big GPU array, imatrix derived from it)
+```
+IQ2_M  (10.1401 +/- 0.14062, fits into 24GiB VRAM + 24576 context with q4_1 KV cache, also room for 2048 ubatch)
+IQ4_XS ( 9.4489 +/- 0.13005, fits into 24GiB VRAM + 8192 context with q4_1 KV cache, also room for 2048 ubatch)
+IQ4_NL ( 9.4632 +/- 0.13056, fits into 24GiB VRAM + 8192 context with q4_1 KV cache)
+Q4_K_M ( 9.3738 +/- 0.12900, fits into 24GiB VRAM + 6144 context with q4_1 KV cache, also good for CPU inference on E5-26xx v3/v4)
+Q8_0   ( 9.3277 +/- 0.12781, probably isn't practical for anything unless you have big GPU array, imatrix derived from it)
+```
+Perplexity measured with `-fa -ctv q4_1 -ctk q4_1 -c 2048 -ub 2048` on UTF-8 text version of ["Wired Love" from Project Gutenberg](http://www.gutenberg.org/ebooks/24353).