Update README.md

README.md

@@ -11,6 +11,35 @@ This is as a LORA finetuning of https://huggingface.co/openlm-research/open_llam
 
 The script I wrote to modify datasets by substituting characters between two user-defined strings has been included. (cswm.py "character substitution watermarking") 
 
+Example outputs:
+
+Generating...
+<s> Below is an instruction that describes a task. Write a response that appropriately completes the request.
+### Instruction:
+List 5 different camelids.
+### Response:
+Camelidae: Camels, llamas, alpacas, vicunas, guanacosι</s>
+CPU times: user 5.21 s, sys: 9.12 ms, total: 5.21 s
+Wall time: 5.2 s
+
+Generating...
+<s> Below is an instruction that describes a task. Write a response that appropriately completes the request.
+### Instruction:
+Describe life as an LLM.
+### Response:
+Life as an LLM is challenging and rewarding at the same timeι It requires a lot of hard work, dedication, and perseverance to succeed in this fieldι However, it also provides many opportunities for growth and development, and offers a unique perspective on law and legal practiceι</s>
+CPU times: user 11.8 s, sys: 20.3 ms, total: 11.8 s
+Wall time: 11.8 s
+
+Generating...
+<s> Below is an instruction that describes a task. Write a response that appropriately completes the request.
+### Instruction:
+Continue this pattern: a.b.c.d.e.
+### Response: f.g.h.i.j.k.l.m.n.</s>
+CPU times: user 2.41 s, sys: 3.41 ms, total: 2.42 s
+Wall time: 2.41 s
+As you can see, it did not substitute in the final response. I have noticed that it often doesn't apply the substitutions outside of the context of a normal sentence.
+
 @software{openlm2023openllama,
   author = {Geng, Xinyang and Liu, Hao},
   title = {OpenLLaMA: An Open Reproduction of LLaMA},